An FBSDE approach to the Skorokhod embedding problem for Gaussian processes with non-linear drift

Size: px
Start display at page:

Download "An FBSDE approach to the Skorokhod embedding problem for Gaussian processes with non-linear drift"

Transcription

1 See icuion, a, an auho pofile fo hi publicaion a: hp:// An FBSDE appoach o he Skookho embeing poblem fo Gauian pocee wih non-linea if Aicle in ELECTRONIC JOURNAL OF PROBABILITY Augu 214 Impac Faco:.77 DOI: /EJP.v Souce: axiv CITATIONS 2 READS 28 3 auho: Alexane Fomm Humbol-Univeiä zu Belin 4 PUBLICATIONS 7 CITATIONS Pee Imkelle Humbol-Univeiä zu Belin 32 PUBLICATIONS 43 CITATIONS SEE PROFILE SEE PROFILE Davi Johanne Pömel ETH Zuich 9 PUBLICATIONS 17 CITATIONS SEE PROFILE Available fom: Davi Johanne Pömel Reieve on: 8 May 216

2 An FBSDE appoach o he Skookho embeing poblem fo Gauian pocee wih non-linea if Alexane Fomm, Pee Imkelle, Davi J. Pömel axiv: v1 [mah.pr 27 Aug 214 Iniu fü Mahemaik Humbol-Univeiä zu Belin Une en Linen Belin Gemany Augu 28, 214 Abac We olve he Skookho embeing poblem fo a cla of Gauian pocee incluing Bownian moion wih non-linea if. Ou appoach elie on olving an aociae ongly couple yem of Fowa Backwa Sochaic Diffeenial Equaion FBSDE, an inveigaing he egulaiy of he obaine oluion. Fo hi pupoe we exen he exience, uniquene an egulaiy heoy of o calle ecoupling fiel fo Makovian FBSDE o a eing in which he coefficien ae only locally Lipchiz coninuou. MSC 21: Pimay: 6G4, 6H3; econay: 93E2. Key wo an phae: BMO poce, BSDE, ecoupling fiel, fowa backwa ochaic iffeenial equaion, FBSDE, Skookho embeing, vaiaional iffeeniaion. 1 Inoucion The Skookho embeing poblem SEP imulae eeach in pobabiliy heoy now fo ove 5yea. Theclaical goal ofhesepconi infining,foagivenbownianmoion W an a pobabiliy meaue ν, a opping ime τ uch ha W τ poee he law ν. I wa fi fomulae an olve by Skookho [Sko61, Sko65 in Since hen hee appeae many iffeen conucion fo he opping ime τ an genealizaion of he oiginal poblem in he lieaue. Ju o name ome of he mo famou oluion o he SEP we efe o Roo [Roo69, Ro [Ro71 an Azéma-Yo [AY79. A compehenive uvey can be foun in [Ob l4. Recenly, he Skookho embeing aie aiional inee becaue of ome applicaion in financial mahemaic, a fo inance o obain moel-inepenen boun on lookback opion [Hob98 o on opion on vaiance [CL1, CW13, OR13. An inoucion o hi cloe connecion of he Skookho embeing poblem an obu financial mahemaic can A.F. an D.J.P. wee financially uppoe by Ph.D. cholahip of he DFG Reeach Taining Goup 1845 Sochaic Analyi wih Applicaion in Biology, Finance an Phyic. 1

3 be foun in [Hob11. In hi pape we conuc a oluion o he Skookho embeing poblem fo Gauian poce G of he fom G := G + α + β W, whee G R i a conan an α,β: [, R ae uiable funcion. Epecially, hi cla of pocee inclue Bownian moion wih non-linea if. The SEP fo Bownian moion wih linea if wa fi olve in he echnical epo [Hal68 an 3 yea lae again in [GF an [Pe. Technique evelope in hee wok can be exene o imehomogeneou iffuion, a one in[pp1, an can be een a genealizaion of he Azéma-Yo oluion. Howeve, o he be of ou knowlege hee exi no oluion o fa fo he cae of a Bownian moion wih non-linea if. The pii of ou appoach i elae o he one by Ba [Ba83, who employe maingale epeenaion o fin an alenaive oluion of he SEP fo he Bownian moion. Thi appoach wa fuhe evelope fo he Bownian moion wih linea if in [AHI8 an fo ime-homogeneou iffuion in [AHS13. I e upon he obevaion ha he SEP may be viewe a he weak veion of a ochaic conol poblem: he goal i o ee G in uch a way ha i ake he iibuion of a pecibe law, which in cae of zeo if i cloely elae o he maingale epeenaion of a anom vaiable wih hi law. We heefoe popoein hi pape o fomulae anolve hesep fo Gin em of afully couple Fowa Backwa Sochaic Diffeenial Equaion FBSDE. In geneal em, he ynamic of a yem of FBSDE i expee by he equaion X = X + Y = ξx T µ,x,y,z + f,x,y,z σ,x,y,z W, Z W, [,T, wih coefficien funcion µ,σ of he fowa pa, eminal coniion ξ an ive f of he backwa componen. In ecen ecae he heoy of FBSDE wih i cloe connecion o quai-linea paial iffeenial equaion an hei vicoiy oluion ha been popagae exenively, in paicula in i numeou aea of applicaion a ochaic conol an mahemaical finance ee [EPQ97 o [PW99. Thee ae mainly hee meho o how he exience of a oluion fo a yem of FBSDE: he conacion meho [An93, PT99, he fou ep cheme [MPY94 an he meho of coninuaion [HP95, Yon97, PW99. A a unifie appoach, [MWZZ11 ee alo [Del2 eigne he heoy of ecoupling fiel fo FBSDE, which wa ignificanly efine in [FI13. I can pimaily be een a an exenion of he conacion meho. In ou appoach of he SEP via FBSDE, we hall focu on he ubcla of Makovian one fo which all involve coefficien funcion ξ,µ,σ,f ae eeminiic. We, howeve, have o allow fo no globally, bu only locally Lipchiz coninuou coefficien µ, σ, f in he conol vaiable z, an heefoe o evelop an exience, uniquene an egulaiy heoy fo FBSDE in hi famewok. Equippe wih hee ool we olve he FBSDE yem euling fom he SEP. We fi conuc a weak oluion, i.e. we obain a Gauian poce of he above fom an an inegable anom ime uch ha, oppe a hi ime, he poce poee he given iibuion ν. Une uiable egulaiy on he given meaue ν an he poce, hi conucion will be caie ove o he oiginally given Gauian poce G. Thi olve he SEP fo G. 2

4 The pape i oganize a follow: in Secion 2 we elae he SEP o a fully couple yem of FBSDE, an in Secion 3 we eablih geneal eul fo ecoupling fiel of FBSDE. The Skookho embeing poblem i olve in Secion 4, in i weak an in i ong veion. Secion A ecall ome auxiliay eul fo BMO pocee. 2 An FBSDE appoach o he Skookho embeing poblem We conie a filee pobabiliy pace Ω,F,F [,,P lage enough o cay a oneimenional Bownian moion W. The filaion F [, i aume o be geneae by he Bownian moion an i aume o be augmene by P-null e. We alo aume ha F = σ = F. We a by fomulaing he Skookho embeing poblem in he moifie veion SEP: Fo given pobabiliy meaue ν on R an a Gauian poce X on [, of he fom X := X + α + β W, 1 whee X R i ome peeemine conan an α,β: [, R ae eeminiic meauable pocee uch ha α + β2 < fo all, fin a F -opping ime τ.. E[τ < ogehe wih a aing poin c R uch ha c+x τ ha he law ν. In oe o have a uly ochaic poblem β houl no vanih an ν houl no be a Diac meaue. In fac we will aume ha β i boune away fom zeo lae on. Ou meho of olving hi poblem i bae on he obevaion ha i may be viewe a he weak veion of a ochaic conol poblem: We wan o ee X in uch a way ha i ake he iibuion of a pecibe law. The pii of ou appoach i elae o an appoach o he oiginal Skookho embeing poblem by Ba [Ba83 ha wa lae exene o he Bownian moion wih linea if in [AHI8. The poceue of boh pape can be biefly ummaize an ivie ino he following fou ep. 1. Conuc a funcion g: R R uch ha gw 1 ha he given law ν. 2. Ue he maingale epeenaion popey of he Bownian moion fo α an β 1 o BSDE echnique fo α κ an β 1 o olve Y = gw 1 κ 1 1 Z 2 Z W, [, Apply he anom ime-change of Dambi, Dubin an Schwaz in he quaaic vaiaion cale. Z2 o anfom he maingale. Z W ino a Bownian moion B. Thi alo povie a anom ime τ := 1 Z2 fulfilling B τ +κ τ +Y = gw 1, which i why B τ +κ τ +Y ha he law ν. 3

5 4. Show ha τ i a opping ime wih epec o he filaion geneae by B hough an explici chaaceizaion uing he unique oluion of an oinay iffeenial equaion. Wih hi ecipion anfom he embeing wih epec o B ino one wih epec o he oiginal Bownian moion W o obain he opping ime τ a he analogue o τ. The fi ep of he algoihm ju keche i faily eay. Le F: R [,1 uch ha Fx := ν, x i he cumulaive iibuion funcion aociae wih ν an efine F 1 :,1 R via F 1 y := inf{x R : Fx y}. DenoingbyΦheiibuionfuncionofheananomaliibuion, weefineg: R R a gx := F 1 Φx. I i aighfowa o pove ha g ha he following popeie. Lemma 1. The funcion g i meauable an non-eceaing. Moeove, if ν i no a Diac meaue, hen g i no ienically conan an gw 1 ha he law ν. Poof. Since Φ an F 1 ae meauable an non-eceaing, hei compoiion g i alo meauable an non-eceaing. Clealy, g can only be conan if F 1 i conan, which can only happen if F aume value in {,1}. Thi only happen in cae ν i a Diac meaue. In oe o ee ha gw 1 ha he law ν, noe ha PgW 1 x = PF 1 ΦW 1 x = PW 1 Φ 1 Fx = ΦΦ 1 Fx = Fx fo all x R. Now efine a meauable funcion ˆδ: [, R via ˆδ := X + α uch ha X = ˆδ+ β W. Obviouly, ˆδ i weakly iffeeniable. Conveely, fo evey weakly iffeeniable funcion ˆδ: [, R we can e X := ˆδ an α := ˆδ. Fuhemoe, efine H: [, [, via H := β 2. Noe ha H i weakly iffeeniable, monoonically inceaing an a a. If we aume ha β i boune away fom, H become icly inceaing an inveible uch ha he invee funcion H 1 i monoonically inceaing an Lipchiz coninuou. In hi cae we can efine δ := ˆδ H 1. If β 1, hen H = I an hu δ = ˆδ. Fo he econ ep we aume ha β i boune away fom an obeve ha he anom ime change, which un he maingale Z W ino a Gauian poce of he fom β B imulaneouly un he cale poce. Z2 ino β2 = H. Thi mean we have o moify he claical maingale epeenaion of gw 1 o 1 [ 1 1 gw 1 + H ˆδ 1 Z 2 E gw 1 + H ˆδ 1 Z 2 = Z W, 4

6 which amoun o fining a oluion Y,Z o he equaion 1 Y = gw 1 δ Z 2 1 Z W, [,1. 3 Fo δ hi woul be ju he uual maingale epeenaion wih epec o he Bownian moion. Alo fo a linea if δ = κ an β 1 equaion 3 can be ewien a Ỹ := Y +κ Z 2 = gw 1 κ 1 Z 2 1 Z W, [,1, which i exacly he BSDE 2 elae o he SEP a ae in [AHI8. In he cae of a Bownian moion wih geneal if equaion 3 woul be a BSDE wih ime-elaye eminal coniion. Unfounaely, he heoy of BSDE wih ime-elay a inouce by Delong an Imkelle in [DI1 an exene by Delong [Del12 fo ime-elaye eminal coniion eache i limi in ou iuaion. Alenaively, we will unean equaion 3 a an FBSDE an evelop new echnique o olve i. Thi will be one in Secion 3 an 4. Befoe we ackle he olvabiliy of equaion 3, we how ha i eally lea o he eie eul in he hi ep of ou algoihm. To be mahemaically igoou we inouce S 2 R a he pace of all pogeively meauable pocee Y : Ω [,1 R aifying up [,1 E[ Y 2 <, H 2 R a he pace of all pogeively meauable pocee Z: Ω [,1 R aifying E[ 1 Z 2 <, whee enoe he Eucliean nom on R. Fo he e of he pape we aume ha β i boune away fom, i.e. inf [, β >. Lemma 2. Suppoe Y,Z S 2 R H 2 R i a oluion of 3. Then hee exi a Bownian moion B an a anom ime τ wih E[ τ < uch ha τ τ Y +X + α + β B = gw 1. Poof. Noe ha Y i a maingale wih quaaic vaiaion poce Z2 fo [,1 ince Z H 2 R. Now chooe anohe Bownian moion B which i inepenen of Y. If neceay we exen ou pobabiliy pace uch ha i accommoae he Bownian moion B. Se τ := H 1 1 Z2, an efine he ime-change of he ype of Dambi, Dubin an Schwaz by { { inf : σ := Z2 > } β2 if < τ 1 if τ. Obeve ha he coniion < τ i equivalen o β2 < 1 Z2. Since Y σ i a coninuou maingale wih quaaic vaiaion H = β2, we can efine a Bownian moion B by τ 1 B := B B τ + Y σ, <. β 5

7 We fin an fuhe τ 1 β B + ˆδ τ+y = Y 1 Y +δ Z 2 +Y = gw 1, 1 E[ τ = E [H 1 Z 2 <, whee we ue ha Z H 2 R an H 1 i Lipchiz coninuou. A an immeiae conequence of he peviou lemma we obeve he following fac. If we have a oluion Y,Z S 2 R H 2 R of equaion 3, we obain a weak oluion o he Skookho embeing poblem, i.e. a Gauian poce of he fom 1, a aing poin c, an an inegable anom ime uch ha ou poce oppe a hi ime poee a given iibuion. A a fi glance equaion 3 migh look eay. We, howeve, have o eal wih a fully couple FBSDE which in aiion poee a no globally Lipchiz coninuou coefficien in he fowa componen. 3 Decoupling fiel fo fully couple FBSDE The heoy of FBSDE, cloely connece o he heoy of quai-linea paial iffeenial equaion an hei vicoiy oluion, eceive i geneal inee fom numeou aea of applicaion among which ochaic conol an mahemaical finance ae he mo vivi one in ecen ecae ee [EPQ97 o [PW99. Owing o hei geneal ignificance, we ea he heoy of FBSDE an hei ecoupling fiel in a moe geneal famewok han migh be neee o obain a oluion o ou equaion 3. Alhough in Secion 3.2 we will focu on he Makovian cae, which mean ha all involve coefficien ae puely eeminiic, le u well in a moe geneal eing fi. 3.1 Geneal ecoupling fiel Fo a fixe ime hoizon T >, we conie a complee filee pobabiliy pace Ω,F,F [,T,P, whee F conain all null e, W [,T i a -imenional Bownian moion inepenen of F, an F := σf,w [, wih F := F T. The ynamic of an FBSDE i claically given by X = X + Y = ξx T µ,x,y,z + f,x,y,z σ,x,y,z W, Z W, fo, [,T an X R n, whee ξ,µ,σ,f ae meauable funcion. Moe peciely, ξ: Ω R n R m, µ: [,T Ω R n R m R m R n, σ: [,T Ω R n R m R m R n, f: [,T Ω R n R m R m R m, 6

8 fo n,m, N. Thoughou he whole ecion µ, σ an f ae aume o be pogeively meauable wih epec o F [,T, i.e. µ1 [,,σ1 [,,f1 [, ae B[,T F BR n BR m BR m -meauable fo all [,T. A ecoupling fiel come wih an even iche ucue han ju a claical oluion. Definiion 3. Le [,T. A funcion u: [,T Ω R n R m wih ut, = ξ a.e. i calle ecoupling fiel fo ξ,µ,σ,f on [,T if fo all 1, 2 [,T wih 1 2 an any F 1 -meauable X 1 : Ω R n hee exi pogeive pocee X,Y,Z on [ 1, 2 uch ha X = X 1 + Y = Y 2 1 µ,x,y,z + 2 f,x,y,z 1 σ,x,y,z W, 2 Z W, Y = u,x, 4 fo all [ 1, 2. In paicula, we wan all inegal o be well-efine an X,Y,Z o have value in R n, R m an R m, epecively. Some emak abou hi efiniion ae in place. The fi equaion in 4 i calle he fowa equaion, he econ he backwa equaion an he hi will be efee o a he ecoupling coniion. The equiemen ha X houl a a X 1 i efee o a he iniial coniion. By a ligh abue of noaion we will omeime efe o X 1 ielf a he iniial coniion. Noe ha, if 2 = T, we ge Y T = ξx T a.. a a conequence of he ecoupling coniion ogehe wih ut, = ξ. A he ame ime Y T = ξx T ogehe wih ecoupling coniion implie ut, = ξ a.e. If 2 = T we can ay ha a iple X,Y,Z olve he FBSDE, meaning ha i aifie he fowa an he backwa equaion, ogehe wih Y T = ξx T. Thi elaionhip Y T = ξx T i efee o a he eminal coniion. By an abueof noaion he funcion ξ ielf i alo omeime efeeo a heeminal coniion. Someime we will ecibe he elaionhip ut, = ξ a.e. wih hi em. In cona o claical oluion of FBSDE, ecoupling fiel on iffeen ineval can be pae ogehe. Lemma 4 Lemma 1 in [FI13. Le u be a ecoupling fiel fo ξ,µ,σ,f on [,T an ũ be a ecoupling fiel fo u,,µ,σ,f on [,, fo < < T. Then, he map û given by û := ũ1 [, +u1,t i a ecoupling fiel fo ξ,µ,σ,f on [,T. We wan o emak ha, if u i a ecoupling fiel an ũ i a moificaion of u, i.e. fo each [,T he funcion u,ω, an ũ,ω, coincie fo almo all ω Ω, hen ũ i alo a ecouplingfielo heamepoblem. Soucoul alo beefeeo a acla of moificaion. Some of he epeenaive of he cla migh be pogeively meauable, ohe no. A we ee below a pogeively meauable epeenaive oe exi if he ecoupling fiel i Lipchiz coninuou in x: 7

9 Lemma 5 Lemma 2 in [FI13. Le u: [,T Ω R n R m be a ecoupling fiel o ξ,µ,σ,f which i Lipchiz coninuou in x R n in he ene ha hee exi a conan L >.. fo evey [,T: u,ω,x u,ω,x L x x x,x R n, fo a.a. ω Ω. Then u ha a moificaion ũ which i pogeively meauable an Lipchiz coninuou in x in he ong ene ũ,ω,x ũ,ω,x L x x [,T, ω Ω, x,x R n. Le I [,T bean ineval an u : I Ω R n R m a map uch ha u, i meauable fo evey I. We efine L u,x := upinf{l fo a.a. ω Ω : u,ω,x u,ω,x L x x fo all x,x R n }, I whee inf :=. We alo e L u,x := if u, i no meauable fo evey I. One can how ha L u,x < i equivalen o u having a moificaion which i uly Lipchiz coninuou in x R n. WeenoebyL σ,z helipchizconan ofσ w... heepenenceonhela componen z an w... he Fobeniu nom on R m an R n. We e L σ,z = if σ i no Lipchiz coninuou in z. By L 1 σ,z = 1 L σ,z we mean 1 L σ,z if L σ,z > an ohewie. Definiion 6. Le u: [,T Ω R n R m be a ecoupling fiel o ξ,µ,σ,f. We ay u o be weakly egula if L u,x < L 1 σ,z an up [,T u,, <. Thi i a naual efiniion ue o Lemma 5. In pacice, howeve, i i impoan o have explici knowlege abou he egulaiy of X, Y, Z. Fo inance, i i impoan o know in which pace he pocee live, an how hey eac o change in he iniial value. Specifically, i can be vey ueful o have iffeeniabiliy of X, Y, Z w... he iniial value. In he following we nee fuhe noaion. Fo an inegable eal value anom vaiable F he expeion E [F efe o E[F F, while Eˆ, [F efe o eupe[f F, which migh be, bu i alway well efine a he infimum of all conan c [, uch ha E[F F c a.. Aiionally, we wie F fo he eenial upemum of F. Definiion 7. Leu: [,T Ω R n R m beaweaklyegulaecouplingfieloξ,µ,σ,f. We call u ongly egula if fo all fixe 1, 2 [,T, 1 2, he pocee X,Y,Z aiing in 4 ae a.e unique an aify [ 2 up E 1, [ X 2 + up E 1, [ Y 2 +E 1, Z 2 <, 5 [ 1, 2 [ 1, 2 1 foeach conan iniialvaluex 1 = x R n. Inaiionheymubemeauableafuncion of x,,ω an even weakly iffeeniable w... x R n uch ha fo evey [ 1, 2 he mapping X an Y ae meauable funcion of x,ω an even weakly iffeeniable w... x uch ha eup x R n up up E 1, [ x X 2 <, v S n 1 [ 1, 2 8 v

10 eup x R n eup x R n up v S n 1 [ 1, 2 up v S n 1 E 1, up E 1, [ x Y 2 <, v [ 2 x Z 2 <. 6 We ay ha a ecoupling fiel on [,T i ongly egula on a ubineval [ 1, 2 [,T if u eice o [ 1, 2 i a ongly egula ecoupling fiel fo u 2,,µ,σ,f. Une ceain coniion a ich exience, uniquene an egulaiy heoy fo ecoupling fiel can be evelope. We will ummaize he main eul, which ae poven in [FI13: Aumpion SLC: ξ,µ, σ, f aifie ana Lipchiz coniion SLC if 1. µ,σ,f ae Lipchiz coninuou in x,y,z wih Lipchiz conan L, 2. µ + f + σ,,,, <, 3. ξ: Ω R n R m i meauable uch ha ξ, < an L ξ,x < L 1 σ,z. Theoem 8 Theoem 1 in [FI13. Suppoe ξ,µ,σ,f aifie SLC. Then hee exi a ime [,T uch ha ξ,µ,σ,f ha a unique up o moificaion ecoupling fiel u on [,T wih L u,x < L 1 σ,z an up [,T u,, <. A bief icuion of exience an uniquene of claical oluion can be foun in Remak 3 in [FI13. Fo lae efeence we give he following emak cf. Remak 1 an 2 in [FI13. Remak 9. I can be obeve fom he poof ha he upemum of all h = T, wih aifying he popeie equie in Theoem 8 can be boune away fom by a boun, which only epen on he Lipchiz conan of µ,σ,f w... he la 3 componen, T, L ξ an L ξ L σ,z < 1, an which i monoonically eceaing in hee value. Remak 1. I can be obeve fom he poof ha ou ecoupling fiel u on [,T aifie L u,,x L ξ,x +CT 1 4, whee C i ome conan which oe no epen on [,T. Moe peciely, C epen only on T, L, L ξ,x, L ξ,x L σ,z an i monoonically inceaing in hee value. We can yemaically exen hi local heoy o obain global eul. Thi i bae on a imple agumen which we will efe o a mall ineval inucion. Lemma 11 Lemma 11 an 12 in [FI13. Le T 1 < T 2 be eal numbe an le S [T 1,T Fowa: If T 1 S an hee exi an h >.. [,+h [T 1,T 2 S fo all S, hen S = [T 1,T 2 an in paicula T 2 S. 2. Backwa: If T 2 S an hee exi an h >.. [ h, [T 1,T 2 S fo all S, hen S = [T 1,T 2 an in paicula T 1 S. Uing hee imple eul we obain global uniquene an global egulaiy of a ecoupling fiel. 1 v 9

11 Theoem 12 Coollay 1 an 2 in [FI13. Suppoe ha ξ,µ,σ,f aifie SLC. 1. Global uniquene: If hee ae wo weakly egula ecoupling fiel u 1,u 2 o he coeponing poblem on ome ineval [,T, hen we have u 1 = u 2 up o moificaion. 2. Global egulaiy: If hee exi a weakly egula ecoupling fiel u o hi poblem on ome ineval [,T, hen u i ongly egula. Noice ha Theoem 12 only povie uniquene of weakly egula ecoupling fiel, no uniquene of pocee X, Y, Z olving he FBSDE in he claical ene. Howeve, uing global egulaiy in Theoem 12 one can how: Coollay 13 Coollay 3 in [FI13. Le ξ,µ,σ,f fulfill SLC. If hee exi a weakly egula ecoupling fiel u of he coeponing FBSDE on ome ineval [,T, hen fo any iniial coniion X = x R n hee i a unique oluion X,Y,Z of he FBSDE on [,T aifying [ up E[ X 2 + up E[ Y 2 +E Z 2 <. [,T [,T 3.2 Makovian ecoupling fiel A yem of FBSDE given by ξ,µ,σ,f i ai o be Makovian if hee fou coefficien funcion ae eeminiic, ha i, if hey epen only on, x, y, z. In he Makovian iuaion we can omewha elax he Lipchiz coninuiy aumpion an ill obain local exience ogehe wih uniquene. Wha make he Makovian cae o pecial i he popey Z = u x,x σ,x,y,z, which come fom he fac ha u will alo be eeminiic. Thi popey allow u o boun Z by a conan if we aume ha σ i boune. Lemma 14 Lemma 14 in [FI13. Le µ,σ,f,ξ aify SLC an aume in aiion ha hey ae eeminiic. Aume ha we have a weakly egula ecoupling fiel u on an ineval [,T. Then u i eeminiic in he ene ha i ha a moificaion which i a funcion of,x [,T R n only. An applicaion of Lemma 14 i he following vey funamenal eul. Lemma 15 Lemma 15 in [FI13. Le ξ,µ, σ, f aify SLC an uppoe ha hee coefficien funcion ae eeminiic. Le u be a weakly egula ecoupling fiel on an ineval [,T. Chooe 1 < 2 fom [,T an an iniial coniion X 1. Then he coeponing Z aifie Z L u,x σ. If Z <, we alo have Z L u,x σ,,, 1 L u,x L σ,z 1. Nex we inveigae he coninuiy of u a a funcion of ime an pace. Lemma 16 Lemma 16 in [FI13. Aume ha µ,σ,f have linea gowh in x,y in he ene µ + σ + f,ω,x,y,z C1+ x + y,x,y,z [,T R n R m R m, 1

12 fo a.a. ω Ω, whee C [, i ome conan. If u i a ongly egula an eeminiic ecoupling fiel o ξ,µ,σ,f on an ineval [,T, hen u i coninuou in he ene ha i ha a moificaion which i a coninuou funcion on [,T R n. Thi bounene of Z in he Makovian cae moivae he following efiniion. I will allow u o evelop a heoy fo non-lipchiz poblem via uncaion. Definiion 17. Le [,T. We call a funcion u: [,T Ω R n R m wih ut,ω, = ξω, fo a.a. ω Ω a Makovian ecoupling fiel fo ξ,µ,σ,f on [,T if fo all 1, 2 [,T wih 1 2 an any F 1 -meauable X 1 : Ω R n hee exi pogeive pocee X,Y,Z on [ 1, 2 uch ha he equaion in 4 hol a.. fo all [ 1, 2, an aiionally Z <. We emak ha a Makovian ecoupling fiel i alway a ecoupling fiel in he ana ene a well. The only iffeence i ha we ae only ineee in iple X,Y,Z, whee Z i a.e. boune. Regulaiy fo Makovian ecoupling fiel i efine vey imilaly o ana egulaiy. Definiion 18. Le u: [,T Ω R n R m be a Makovian ecoupling fiel o ξ,µ,σ,f. We call u weakly egula if L u,x < L 1 σ,z an up [,T u,, <. We call a weakly egula u ongly egula if fo all fixe 1, 2 [,T, 1 2, he pocee X, Y, Z aiing in he efining popey of a Makovian ecoupling fiel ae a.e. unique fo each conan iniial value X 1 = x R n an aify 5. In aiion hey mu be meauable a funcion of x,, ω an even weakly iffeeniable w... x R n uch ha fo evey [ 1, 2 he mapping X an Y ae meauable funcion of x,ω, an even weakly iffeeniable w... x uch ha 6 hol. We ay ha a Makovian ecoupling fiel on [,T i ongly egula on a ubineval [ 1, 2 [,T if u eice o [ 1, 2 i a ongly egula Makovian ecoupling fiel fo u 2,,µ,σ,f. Now we efine a cla of poblem fo which an exience an uniquene heoy will be evelope. Aumpion MLLC: ξ,µ, σ, f fulfill a moifie local Lipchiz coniion MLLC if 1. he funcion µ,σ,f ae a eeminiic, b Lipchiz coninuou in x,y,z on e of he fom [,T R n R m B, whee B R m i an abiay boune e, c an fulfill µ,,,, f,,,, σ,,,,l σ,z <, 2. ξ: R n R m aifie L ξ,x < L 1 σ,z. We a a poviing a local exience eul. 11

13 Theoem 19. Le ξ,µ,σ,f aify MLLC. Then hee exi a ime [,T uch ha ξ,µ,σ,f ha a unique weakly egula Makovian ecoupling fiel u on [,T. Thi u i alo ongly egula, eeminiic, coninuou an aifie up 1, 2,X 1 Z <, whee 1 < 2 ae fom [,T an X 1 i an iniial value ee he efiniion of a Makovian ecoupling fiel fo he meaning of hee vaiable. Poof. Fo any conan H > le χ H : R m R m be efine a χ H z := 1 { z <H} z + H z 1 { z H}z. I i eay o check ha χ H i Lipchiz coninuou wih Lipchiz conan L χh = 1 an boune by H. Fuhemoe, we have χ H z = z if z H. We implemen an inne cuoff by efining µ H,σ H,f H via µ H,x,y,z := µ,x,y,χ H z, ec. The bounene of χ H ogehe wih i Lipchiz coninuiy make µ H,σ H,f H Lipchiz coninuou wih ome Lipchiz conan L H. Fuhemoe, L σh,z L σ,z. Alo µ H,σ H,f H have linea gowh in y,z a equie by Lemma 16. Accoing o Theoem 8 we know ha he poblem given by ξ,µ H,σ H,f H ha a unique weakly egula ecoupling fiel u on ome mall ineval [,T whee [,T. We alo know ha hi u i ongly egula, u i eeminiic by Lemma 14, an coninuou by Lemma 16. We will how ha fo ufficienly lage H an [,T i will alo beamakovian ecoupling fiel o he poblem ξ,µ,σ,f. Uing Remak 1 L u,,x L ξ,x +C H T 1 4 [,T, whee C H < i a conan which oe no epen on [,T. Fo any 1 [,T an F 1 - meauable iniial value X 1 conie he coeponing unique X,Y,Z on [ 1,T aifying he fowa equaion, he backwa equaion an he ecoupling coniion fo µ H,σ H,f H an u. Uing Lemma 15 we have Z L u,x σ H L u,x σ,,, +L σ,z H < an, heefoe, Z up [ 1,TL u,,x σ,,, 1 up [1,TL u,,x L σ,z L ξ,x +C H T σ,,, 1 L ξ,x L σ,z L σ,z C H T = L ξ,x σ,,, 1 L ξ,x L σ,z L σ,z C H T C HT σ,,, 1 L ξ,x L σ,z L σ,z C H T fo T 1 mall enough. Now we only nee o chooe H lage enough uch ha L ξ,x σ,,, 1 L ξ,x L σ,z become malle han H 4, an hen in he econ ep chooe cloe enough o T, uch ha L σ,z C H T 1 4 become malle han L ξ,xl σ,z, C H σ,,, T L ξ,x L σ,z become malle han H 4. 12

14 Conieing 7 hi implie ha if 1 [,T he poce Z a.e. oe no leave he egion in which he cuoff i paive, i.e. he ball of aiu H. Theefoe, u eice o he ineval [,T i a ecoupling fiel o ξ,µ,σ,f, no ju o ξ,µ H,σ H,f H. I i even a Makovian ecoupling fiel ue o he bounene of Z. A a Makovian ecoupling fiel i i weakly egula, becaue i i weakly egula a a ecoupling fiel o ξ,µ H,σ H,f H. Uniquene: Aume han hee i anohe weakly egula Makovian ecoupling fiel ũ o ξ,µ,σ,fon [,T. Chooe a 1 [,T an an x R n a iniial coniion X 1 = x, an conie he coeponing pocee X,Ỹ, Z ha aify he coeponing FBSDE on [ 1,T, ogehe wih he ecoupling coniion via ũ. A he ame ime conie X,Y,Z olving he ame FBSDE on [ 1,T, bu aociae wih he Makovian ecoupling fiel u. Since Z,Z ae boune, he wo iple X,Ỹ, Z an X,Y,Z alo olve he Lipchiz FBSDE given by ξ,µ H,σ H,f H on [ 1,T fo H lage enough. The wo coniion Ỹ = ũ, X an Y = u,x imply by Remak 3 in [FI13 ha boh iple ae pogeively meauable pocee on [ 1,T Ω.. [ up E, X 2 [ + up E, Y 2 [ +E, Z 2 [ 1,T [ 1,T 1 < an coincie. In paicula, ũ 1,x = Ỹ 1 = Y 1 = u 1,x. Song egulaiy of u a a Makovian ecoupling fiel o ξ,µ,σ,f follow iecly fom he above agumen abou uniquene of X, Y, Z fo eeminiic iniial value an boune Z, an he ong egulaiy of u a ecoupling fiel o ξ,µ H,σ H,f H. Remak 2. We obeve fom he poof ha he upemum of all h = T wih aifying he hypohee of Theoem 19 can be boune away fom by a boun, which only epen on L ξ,x, L ξ,x L σ,z, σ,,,, T, L σ,z, he value L H H [, whee L H i he Lipchiz conan of µ,σ,f on [,T R n R m B H w... o he la 3 componen, whee B H R m enoe he ball of aiu H wih cene, an which i monoonically eceaing in hee value. The following naual concep inouce a ype of Makovian ecoupling fiel fo non- Lipchiz poblem non-lipchiz in z, o which nevehele ana Lipchiz eul can be applie. Definiion 21. Le u be a Makovian ecoupling fiel fo ξ,µ,σ,f. We call u conolle in z if hee exi a conan C > uch ha fo all 1, 2 [,T, 1 2, an all iniial value X 1, he coeponing pocee X,Y,Z fom he efiniion of a Makovian ecoupling fiel aify Z ω C, fo almo all,ω [,T Ω. If fo a fixe iple 1, 2,X 1 hee ae iffeen choice fo X,Y,Z, hen all of hem ae uppoe o aify he above conol. We ay ha a Makovian ecoupling fiel on [,T i conolle in z on a ubineval [ 1, 2 [,T if u eice o [ 1, 2 i a Makovian ecoupling fiel fo u 2,,µ,σ,f ha i conolle in z. 13

15 A Makovian ecoupling fiel u on an ineval,t i ai o be conolle in z if i i conolle in z on evey compac ubineval [,T,T wih C poibly epening on. Remak 22. Ou Makovian ecoupling fiel fom Theoem 19 i obviouly conolle in z: conie 7 ogehe wih he choice of 1 mae in he poof. Remak 23. Le ξ,µ,σ,f aify MLLC, an aume ha we have a Makovian ecoupling fiel u on ome ineval [,T, which i weakly egula an conolle in z. Then u i alo a oluion o a Lipchiz poblem obaine hough a cuoff a in Theoem 19. A uch i i ongly egula Theoem 12 an eeminiic Lemma 14. Bu now Lemma 16 i alo applicable, ince ue o he ue of a cuoff we can aume he ype of linea gowh equie hee. So u i alo coninuou. Lemma 24. Le ξ,µ,σ,f aify MLLC. Fo < < T le u be a weakly egula Makovian ecoupling fiel fo ξ,µ,σ,f on [,T. If u i conolle in z on [, an T i mall enough a equie in Theoem 19 ep. Remak 2, hen u i conolle in z on [,T. Poof. Clealy, u i no ju conolle in z on [,, bu alo on [,T wih a poibly iffeen conan, accoing o Remak 22. Define C a he maximum of hee wo conan. We only nee o conol Z by C fo he cae 1 2 T, he ohe wo cae being ivial. Fo hi pupoe conie he pocee X,Y,Z on he ineval [ 1, 2 coeponing o ome iniial value X 1 an fulfilling he fowa equaion, he backwa equaion an he ecoupling coniion. Since he eicion of hee pocee o [ 1, ill fulfill hee hee popeie we obain Z ω C fo almo all [ 1,, ω Ω. A he ame ime, if we eic X,Y,Z o [, 2, we obeve ha hee eicion aify he fowa equaion, he backwa equaion an he ecoupling coniion fo he ineval [, 2 wih X a iniial value. Theefoe Z ω C hol fo a.a. [, 2, ω Ω a well. The following impoan eul allow u o connec he MLLC-cae o SLC. Theoem 25. Le ξ,µ,σ,f be uch ha MLLC i aifie an aume ha hee exi a weakly egula Makovian ecoupling fiel u o hi poblem on ome ineval [,T. Then u i conolle in z. Poof. Le S [,T be he e of all ime [,T,.. u i conolle in z on [,. Clealy S: Fo he ineval [, = {} one can only chooe 1 = 2 = an o Z: [, Ω R m i P-a.e., inepenenly of he iniial value X 1. So we can ake fo C any poiive value. Le S be abiay. Accoing o Lemma 24 hee exi an h >.. u i conolle in z on [,+h T ince u+h T, < an L u+h T, < L 1 σ,z. Conieing Remak 2 an he equiemen u <, L u,x < L 1 σ,z, we can chooe h inepenenly of. Thi how S = [,T by mall ineval inucion. 14

16 Noe ha Theoem 25 implie ogehe wih Remak 23 ha a weakly egula Makovian ecoupling fiel o an MLLC poblem i eeminiic an coninuou. Such a u will be a ana ecoupling fiel o an SLC poblem if we uncae µ,σ,f appopiaely. We can heeby exen he whole heoy o MLLC poblem: Theoem 26. Le ξ,µ,σ,f aify MLLC. 1. Global uniquene: If hee ae wo weakly egula Makovian ecoupling fiel u 1,u 2 o hi poblem on ome ineval [,T, hen u 1 = u Global egulaiy: If ha hee exi a weakly egula Makovian ecoupling fiel u o hi poblem on ome ineval [,T, hen u i ongly egula. Poof. 1. We know ha u 1 an u 2 ae conolle in z. Chooe a paive cuoff ee poof of Theoem 19 an apply 1. of Theoem u i conolle in z. Chooe a paive cuoff ee poof of Theoem 19 an apply 2. of Theoem 12. Lemma 27. Le ξ,µ,σ,f aify MLLC an aume ha hee exi a weakly egula Makovian ecoupling fiel u of he coeponing FBSDE on ome ineval [, T. Then fo any iniial coniion X = x R n hee i a unique oluion X,Y,Z of he FBSDE on [,T uch ha up E[ X 2 + up E[ Y 2 + Z <. [,T [,T Poof. Exience follow fom he fac ha u i alo ongly egula accoing o 2. of Theoem 26 an conolle in z accoing o Theoem 25. Uniquene follow fomcoollay 13: Aumeheeae wo oluion X,Y,Z an X,Ỹ, Z o he FBSDE on [,T boh aifying he afoemenione boun. Bu hen hey boh olve anslc-confom FBSDE obaine hough a paive cuoff. So hey mu coincie accoing o Coollay 13. Definiion 28. Le I M max [,T fo ξ,µ,σ,f be he union of all ineval [,T [,T uch ha hee exi a weakly egula Makovian ecoupling fiel u on [,T. Unfounaely, he maximal ineval migh vey well be open o he lef. Theefoe, we nee o make ou noion moe pecie in he following efiniion. Definiion 29. Le < T. We call a funcion u:,t R n R m a Makovian ecoupling fiel fo ξ,µ,σ,f on,t if u eice o [,T i a Makovian ecoupling fiel fo all,t. A Makovian ecoupling fiel u on,t i ai o be weakly egula if u eice o [,T i a weakly egula Makovian ecoupling fiel fo all,t. A Makovian ecoupling fiel u on,t i ai o be ongly egula if u eice o [,T i ongly egula fo all,t. Theoem 3 Global exience in weak fom. Le ξ,µ,σ,f aify MLLC. Then hee exi a unique weakly egula Makovian ecoupling fiel u on I M max. Thi u i alo eeminiic, conolle in z an ongly egula. Moeove, eihe I M max = [,T o IM max = M min,t, whee M min < T. 15

17 Poof. Le Imax. M Obviouly, hee exi a Makovian ecoupling fiel ǔ on [,T aifying Lǔ,x < L 1 σ,z an up [,T ǔ,, <. ǔ i conolle in z an ongly egula ue o Theoem 25 an 26. We can fuhe aume w.l.o.g. ha ǔ i a coninuou funcion on [,T R n accoing o Remak 23. Thee i only one uch ǔ accoing o Theoem 26. Fuhemoe, fo, Imax M he funcion ǔ an ǔ coincie on [,T becaue of Theoem 26. Define u, := ǔ, fo all Imax. M Thi funcion u i a Makovian ecoupling fiel on [,T, ince i coincie wih ǔ on [,T. Theefoe, u i a Makovian ecoupling fiel on he whole ineval Imax M an aifie L u [,T,x < L 1 σ,z, up [,T u [,T,, < fo all Imax. M Uniquene of u follow iecly fom Theoem 26 applie o evey ineval [,T Imax. M Aeing he fom of Imax M, we ee ha IM max = [,T wih,t i no poible: Aume ohewie. Accoing o he above hee exi a Makovian ecoupling fiel u on [,T.. L u,x < L 1 σ,z an up [,T u,, <. Bu hen u can be exene a lile bi o he lef uing Theoem 19 an Lemma 4, heeby conaicing he efiniion of Imax M. Thefollowing eulbaically ae ha foaingulaiy M min o occuu x ha o exploe a M min. Lemma 31. Le ξ,µ,σ,f aify MLLC. If Imax M = M min,t, hen lim L u,,x = L 1 σ,z, M min whee u i he Makovian ecoupling fiel accoing o Theoem 3. Poof. We ague iniecly. Aume ohewie. Then we can elec ime n M min, n uch ha upl un,,x < L 1 σ,z. n N Bu hen we may chooe an h > accoing o Remak 2 which oe no epen on n an hen chooe n lage enough o have n M min < h. So u can be exene o he lef o a lage ineval [ n h,t conaicing he efiniion of Imax. M 4 Soluion o he Skookho embeing poblem In hi ecion we peen a oluion o he Skookho embeing poblem a ae in SEP a he beginning of Secion 2 bae on oluion of he aociae yem of FBSDE. 4.1 Weak oluion Le u heefoe eun o ou FBSDE 3 ha can be ewien lighly moe geneally a X 1 = x 1 + 1W, X 2 = x 2 + Z 2, Y = gx 1 T δx2 T Z W, u,x 1,X 2 = Y, 8 16

18 fo [,T an x = x 1,x 2 R 2. So uing he noaion of Secion 3 we have µ,ω,x,y,z =,z 2, σ,ω,x,y,z = 1,, f,ω,x,y,z =, ξω,x = gx 1 δx 2, fo all,ω,x,y,z [,T Ω R 2 R R an = 1, n = 2 an m = 1. In paicula, he poblem aifie MLLC. Noice ha by chooing x := x 1,x 2 :=, an T = 1 we will have X 1 1 = W 1 an X 2 1 = 1 Z2, which make he FBSDE equivalen o 3. WihhegenealeulofSecion3.2ahanweaecapableoolvehiyemofequaion. In ohe wo, we ae able o pefom he econ ep of ou algoihm o olve he SEP. Lemma 32. Aume δ an g ae Lipchiz coninuou. Then fo he FBSDE 8 hee exi a unique weakly egula Makovian ecoupling fiel u on [,T. Thi u i ongly egula, conolle in z, eeminiic an coninuou. In paicula, equaion 3 ha a unique oluion Y,Z uch ha Z <. Poof. Uing Theoem 3 we know ha hee exi a unique weakly egula Makovian ecoupling fiel u on Imax. M Thi u i fuhemoe ongly egula, conolle in z, eeminiic an coninuou. I emain o pove Imax M = [,T. Due o Lemma 31 i i ufficien o how he exience of a conan C [, uch ha L u,,x C < L 1 σ,z fo all Imax. M In ou cae L 1 σ,z =, o we have o pove ha he weak paial eivaive of u wih epec o x1 an x 2 ae boh unifomly boune. Fix Imax M an conie he coeponing FBSDE on [,T: Fi noice ha he aociae iple X,Y,Z epen on he iniial value x = x 1,x 2 R 2, even in a weakly iffeeniable way wih epec o he iniial value x, accoing o he ong egulaiy of u. Fo moe on ule egaing woking wih weak eivaive conul Secion 2.2 of [FI13. Le u fi look a he maix xx. We have x 1X1 = 1, x 2X1 =, x 1X2 = 2Z x 2X2 = 1+ x 1Z, 2Z x 2Z, a.. fo [,T, fo almo all x = x 1,x 2 R 2. In paicula, he 2 2-maix x X i inveible if an only if X 2 x 2 i no. We will ee lae ha i emain poiive on he whole ineval allowing u o apply he chain ule of Lemma 5 in oe o wie x u,x x X. Bu le u fi pocee by iffeeniaing he backwa equaion in 8 wih epec o x 2 : x 2Y = δ X 2 T T x 2X2 T x 2Z W. To be pecie he above hol a.. fo evey [,T, fo almo all x = x 1,x 2 R 2. Now efine a opping ime τ via { } τ := inf [,T : x 2X2 T. 17

19 Fo [,τ we have x u,x x X accoing o he chain ule of Lemma 5 an in paicula u,x 1 x 2,X 2 X 2 x 2 = Y x 2. Le u e V := x 2u,X1,X 2, [,T an Z := Z x 2 X 2 x 2 1 { [,τ}. Then he ynamic of X 2 1 x 2 can be expee by 1 τ 1 x τ 2X2 = 1 2Z Z x 2X2, 9 fo an abiay opping ime τ < τ wih value in [,T. We alo have Y x 2 = V an heefoe V = x 2 Y X 2 x 2, [,τ. X 2 x 2 Applying Iô fomula an uing he ynamic of Y an X 2 we eaily obain an x 2 x 2 equaion ecibing he ynamic of V τ : τ 1 τ 1 V τ = V + 2Z Z x 2X2 x 2Y + x 2Z x 2X2 W = V + τ 2Z V Z + τ Z W 1 fo any opping ime τ < τ wih value in [,T. Noe ha, ince V an 2ZV ae boune pocee, Z1[ τ i in BMOP accoing o Theoem 49 wih a BMOP-nom which oe no epen on τ < τ, an o in paicula E[ τ 2Z Z 2 <. Fom 9 we can acually euce ha τ = T mu hol almo uely. Inee, 9 implie ha 1 τ x τ 2X2 = exp 2Z Z o equivalenly τ x 2X2 τ = exp 2Z Z fo all opping ime τ < τ wih value in [,T. Uing coninuiy of whichgive uτ = T a.. becaue{τ < T} τ x 2X2 τ = exp 2Z Z >, { } X 2 x 2 τ = x 2 X 2 we obain, ueoconinuiy of x 2 X 2. So we have X 2 i poiive on he whole [,T an heefoe x 2 xx i inveible on [,T. Seing W := W 2Z V, [,T we can efomulae 1 o V = V + 18 Z W.

20 Thi mean ha V can be viewe a he coniional expecaion of V T = x 2uT,X1 T,X2 T = δ X 2 T wih epec o F an ome pobabiliy meaue, which un W ino a Bownian moion on [,T. Noeheeha 2Z V ibouneon[,t becaue Z <. Hence, weconclueha V an heefoe u,x 1,x 2 i boune by δ x 2 fo almo all x = x 1,x 2 R 2. Thi value i inepenen of. Seconly, we have o boun u,x 1,x 2. To hi en we iffeeniae he equaion in x 1 8 wih epec o x 1 : an efine Noe ha x 1X1 = 1, x 1X2 = 2Z x 1Z, x 1Y = g X 1 T δ X 2 T T x 1X2 T x 1Z W, x 1u,X1,X 2 + x 2u,X1,X 2 x 1X2 = x 1Y, U := x 1u,X1,X 2, Ž := x 1Z Z x 1X2. x 1X2 = U = x 1Y V x 1X2, 2Z Ž + Z x 1X2, which allow u o euce he ynamic of U fom he ynamic of Y, X 2 an V x 1 x 1 uing Iô fomula: U =U + 1 x 1Y V x 1X2 x 1X2 V =U + x 1Z W 2 V Z Ž + Z 2Z V Z + Z W x 1X2 x 1X2 whee he make em eihe mege ino one o cancel ou an he equaion implifie o U =U + =U + 2Z V Ž + Ž W. Ž W. 11 By heame agumen a fo he poce V weeuce ha U an heefoe x 1 u,x 1,x 2 i boune by g = L g fo almo all x 1,x 2, whee L g i he Lipchiz conan of g, 19

21 i.e. he infimum of all Lipchiz conan. Thi how ha Imax M = [,T. Finally, Lemma 27 how ha hee i a unique oluion X,Y,Z o he FBSDE on [,T fo any iniial value X 1,X2 = x 1,x 2 R 2 uch ha up E[ X 2 + up E[ Y 2 + Z <, [,T [,T which i equivalen o he imple coniion Z < a we claim: If Z <, hen accoing o he fowa equaion X 2 x 2 +T Z 2 <, up E[ X 2 = x up E[ W 2 = x 1 2 +T <, [,T [,T an accoing o he backwa equaion ogehe wih he Minkowki inequaliy 1 [ 1 up E[ Y 2 2 = up E[ E gx T δx2 T F [,T [,T E [ gx 1 T δx2 T E g +L g E [ X 1 T whee L g,l δ ae Lipchiz conan of g,δ δ +Lδ E [ gx 1 T 2 1 [ 2 δx 2 + E T [ X 2 T 21 2 <, Fo he following eul we ue he noaion of Secion 2. A befoe we aume ha β i boune away fom. Une hi coniion H 1 i well efine an Lipchiz coninuou. Theefoe, δ = ˆδ H 1 i Lipchiz coninuou if ˆδ i Lipchiz coninuou, which i equivalen o α being boune. Lemma 33. Suppoe g an δ ae boh Lipchiz coninuou wih Lipchiz conan L g an L δ. Then hee exi a Bownian moion B, a anom ime τ H 1 L 2 g an a conan c R uch ha c+ τ α + τ β B ha law ν. Poof. Fi we olve FBSDE 3 uing Lemma 32 uch ha he coeponing Z i boune. Accoing o Lemma 36, which we pove a bi lae, we can even aume ha Z i boune by L g. Now we e c := Y an conuc B an τ a in he poof of Lemma 2. Moeove, τ = H 1 1 Z2 i boune by H 1 L 2 g ince Z i boune by L g an H 1 i inceaing. Remak 34. I i a pioi no clea ha he anom ime τ i alo a opping ime wih epec o F B [, := σb, [, [, a alo menione in Remak 1.2 in [AHI8. Howeve, we will pove a ufficien cieion fo hi in em of egulaiy popeie of he Makovian ecoupling fiel u. Remak 35. The bounene of he opping ime olving he Skookho embeing poblem ha no been inveigae o fequenly. Howeve, vey ecenly i gaine aenion in [AS11 an [AHS13. Epecially, i economic inee come fom i applicaion in he conex of game heoy ee [SS9. 2

22 4.2 Song oluion Thi ubecion i evoe o he fouh ep of ou algoihm, i.e. o anlae he eul of he peceing ecion ino a oluion of he Skookho embeing poblem in he ong ene. Ou main goal i o how ha if g,δ ae ufficienly mooh, hen τ an B conuce o fa will have he popey ha τ i inee a opping ime w... filaion F B [, geneae byhebownian moion B, anhu afuncional of heajecoie of B. Theame funcional applie o he ajecoie of he oiginal Bownian moion W will hen povie he equie ong oluion. Fo hi pupoe, we will aume ha g an δ ae hee ime weakly iffeeniable wih boune eivaive. We will alo equie ha g i non-eceaing an no conan. Ou agumen hall be bae on a eep analyi of egulaiy popeie of he aociae ecoupling fiel u. Fi le u fi pove he following vey ueful eul abou he oluion Y,Z o FBSDE 3 conuce in Lemma 32. Lemma 36. Aume δ an g ae Lipchiz coninuou. Le u be he unique weakly egula Makovian ecoupling fiel aociae o he poblem 8 on [, T conuce in Lemma 32. Then fo any [,T an iniial coniion X 1,X 2 = x 1,x 2 R 2 he aociae poce Z on [,T aifie Z L g = g. Fuhemoe, if he weak eivaive u ha a veion which i coninuou in he fi wo x 1 componen,x 1 on [,T R 2 hen Z ω = x 1u,X 1 ω,x 2 ω fo almo all,ω [,T Ω. Poof. We aleay know ha Z i boune accoing o Lemma 32, bu no in he fom of he moe explici boun Z L g. 1 +h Noice ha lim h h Z ω = Z ω fo almo all ω, Ω [,T ue o he funamenal heoem of Lebegue inegal calculu. 1 +h Now ake ome [,T.. lim h h Z = Z almo uely. Almo all [,T have hi popey. Chooe any h >. +h < T an conie he expeion 1 h E[Y +hw +h W F fo mall h >. On he one han we can wie uing Iô fomula which lea o Y +h W +h W = +h Y W + 1 h E[Y +hw +h W F = 1 h E +h [ +h W W Z W + On he ohe han we can ue he ecoupling coniion o wie Y +h W +h W =u +h,x 1 +h,x2 +h W +h W =u +h,x 1 +h,x2 W +h W + u +h,x 1 +h,x2 +h u +h Z, Z F Z fo h. +h,x 1 +h,x2 W +h W. 21

23 Afe applying coniional expecaion o boh ie of he above equaion we inveigae he wo umman on he igh han ie epaaely. Fi umman: Recall: X 1 an X 2 ae F -meauable, X 1 +h = X1 +W +h W, W +h W i inepenen of F, u i eeminiic, i.e. can be aume o be a funcion of,x 1,x 2 [,T R R only. Thee popeie imply [ E u +h,x 1 +h,x2 W +h W F = u +h,x 1 +z h,x 2 z h 1 e 1 2 z2 z R 2π = x 1u +h,x 1 +z h,x 2 h 1 e 1 2 z2 z, 2π which mean 1 [ lim h h E u +h,x 1 +h,x2 W +h W F R = x 1u,X 1,X 2, if u i coninuou in he fi wo componen on [,T R 2. Hee we ue ha u i x 1 x 1 boune by g accoing o he poof of Lemma 32. Bu even if u i no coninuou x 1 in he fi wo componen, we can ill a lea conol he value by g. 1 [ h E Secon umman: Recall: u +h,x 1 +h,x2 W +h W F u i alo Lipchiz coninuou in he la componen an δ eve a a Lipchiz conan, X 2 +h = X2 + +h Z 2. Thee popeie allow u o eimae 1 [ E u +h,x 1 h +h,x2 +h 1 [ u h E +h,x 1 +h,x2 +h 1 [ +h h E δ Z 2 which clealy en o a h. u u +h,x 1 +h,x2 +h,x 1 +h,x2 W +h W F W+h W F W +h W F 1 h δ h Z 2 E[ W +h W, 22

24 Concluion: We have hown 1 Z = lim h h E[Y 1 [ +hw +h W F = lim h h E u +h,x 1 +h,x2 W +h W F, which i ienical wih x 1 u,x 1,X 2 componen on [,T R 2 an boune by g ohewie. a.. if x 1 u i coninuou in he fi wo FoheequelleubeheuniqueweaklyegulaMakovianecouplingfielohepoblem 8 conuce in Lemma 32. A lea fo he following eul we aume fo convenience T = 1. We alo ue efiniion an noaion fom he poof of Lemma 2. Theoem 37. Aume ha x 1 u i Lipchiz coninuou in he fi wo componen on compac ube of [,1 R 2, R\{} - value on [,1 R 2. Then τ i a opping ime wih epec o he filaion F B = F B [,. Poof. We conie he yem 8 fo = an x 1 = x 2 =. Accoing o Lemma 36 we can aume Z = x 1u,X 1,X 2 an, heeby, have X 2 = Z 2 = fo all [,T. So, we can aume ha X 1 2 x 1u,X 1,X 2 i Lipchiz coninuou an icly inceaing in ue o poiiviy of [,1 R 2, 2 u x on 1 a in. Theefoe, fo evey ω Ω he pah H 1 X 2 ω : [,1 [, i alo Lipchiz coninuou an icly inceaing in ime an, heefoe, ha a coninuou an icly inceaing invee funcion on he ineval [,H 1 X 2 1 ω = [, τω. I i aighfowa o ee ha hi invee i given by he poce σ fom he poof of Lemma 2. We can now calculae he weak eivaive of σ: Fily, noe H 1 x = 1 H H 1 x an 23

25 alo H 1 X 2 σ ω = o equivalenly X 2 σ ω = H. So, we can calculae σ = 1 H 1 X 2 =σ = = 1 H 1 X 2 σ Z 2 σ H u σ x 1,X σ 1,X σ 2 2 = on {σ < 1}. Obeve a hi poin ha { {σ < 1} = < H 1 } X 2 1 = { < τ}. If we efine σ := 1 fo > τ, hen σ i ill coninuou an we have τ = inf{ [, σ 1}. x 1 u β 2 2σ,W σ,h I i alo aighfowa o ee Z σ = uσ x 1,W σ,h fo [, τ. Now, emembe B = 1 β Y σ fo [, τ an alo Y Y = Z W fo [,1, o β Z σ B = β Z σ 1 β Y σ = So, if we efine Σ := W σ, we have ynamic Σ = β uσ x 1,Σ,H B, fo [, τ. So, o um up σ,σ fulfill on [, τ he ynamic σ = + β 2 u x 1 Σ = + + 2σ +,Σ,H β 1 Z σ Z σ W σ = W σ. uσ x 1,Σ,H B, B, whee [, τ. Noe ha hi ynamical yem i locally Lipchiz coninuou in σ,σ. Now, fo any K 1,K 2 > an K 3,1 efine a boune anom vaiable τ K1,K 2,K 3 via τ K1,K 2,K 3 := K 1 inf{ [, Σ K 2 } inf{ [, σ K 3 }. Noe ha σ an Σ boh emain boune on [,τ K1,K 2,K 3. Theefoe, on [,τ K1,K 2,K 3 he pai σ,σ coincie wih he unique oluion σ K 1,K 2,K 3,Σ K 1,K 2,K 3 o a Lipchiz poblem, which i auomaically pogeively meauable w... he filaion F B. Noe ha τ K1,K 2,K 3 = K 1 inf { [, Σ K 1,K 2,K 3 K 2 } inf { [, σ K 1,K 2,K 3 K 3 }, which i clealy a opping ime w... F B. Fuhemoe, ue o coninuiy of Σ an σ τ = up τ K1,K 2,K 3, K 3,1,K 1,K 2 > which make i a opping ime w... F B a well

26 In oe o euce ufficien coniion fo Theoem 37 o hol we nee o inveigae highe oe eivaive of u. Aume ha g, δ, g an δ ae Lipchiz coninuou, an conie he following MLLC yem wih = 1, n = 2 an m = 3: X 1 = x 1 + 1W, X 2 = x 2 + Z 2, Y = gx 1 T δx2 T Z W, u,x 1,X 2 = Y, Y 1 = g X 1 T Y 2 = δ X 2 T Z 1 W Z 2 W 2Z Y 2 Z 1, u 1,X 1,X 2 = Y 1, 2Z Y 2 Z 2, u 2,X 1,X 2 = Y 2. Theoem 38. Fo he above poblem 13 we have Imax M = [,T. Fuhemoe, u = u, u 1 = x 1u an u2 = x 2u, a.e., whee u i he unique weakly egula Makovian ecoupling fiel o he poblem 8. In paicula, u i wice weakly iffeeniable w... x wih unifomly boune eivaive. Poof. The poof i in pa akin o he poof of Lemma 32 an we will eek o keep hee pa ho. Le u i, i =,1,2 be he unique weakly egula Makovian ecoupling fiel on Imax M. We can aume u i o be coninuou funcion on Imax M R2 Theoem 3. Le Imax M. Fo an abiay iniial coniion x R2 conie he coeponing pocee X 1,X 2,Y,Y 1,Y 2,Z,Z 1,Z 2 on [,T. Noe ha X 1,X 2,Y,Z olve FBSDE 8, which implie ha hey coincie wih he pocee X 1,X 2,Y,Z fom 8 if we aume 2 i=1 up [,T E, [ X i 2 + up E, [ Y 2 + Z + [,T 2 i= up [,T E, [ Y i Z i <, accoing o Lemma 27. Thi coniion i fulfille ue o ong egulaiy an he fac ha we wok wih Makovian ecoupling fiel. Now, Y = Y implie u,x = u,x fo all Imax M,x R2, whee Imax M i he maximal ineval fo he poblem given by 13. We now claim ha Y 1,Y 2 ae boune pocee: Uing he backwa equaion we have an, heefoe, [ [ Y 2 = E δ X 2 T E Y 2 δ + 2Z Y 2 Z 2 2 Z Z 2 E [ Y 2 25, i=

27 fo [,T, which uing Gonwall lemma implie [ Y 2 Y 2 = E δ exp 2T Z Z 2. Thi in un auomaically implie bounene of Y 1 accoing o i ynamic. Fuhemoe, Y 1,Z 1 an Y 2,Z 2 aify he BSDE which i alo fulfille by he pocee U,Ž an V, Z fom he poof of Lemma 32 ee 1 an 11 an o in paicula Y 2 V = = Z 2 Z W Z 2 Z W 2Z 2Z Y 2 Z 2 V Z Y 2 V Z 2 +V Z 2 Z. Uing he bounene of Z, Z 2 an V hi implie uing Lemma 48 ha Y 2 V i almo eveywhee. Theefoe, afe eing W := W 2Z V, [,T we ge fom he above equaion T Z 2 Z W = a.. fo [,T. Since W i a Bownian moion une ome pobabiliy meaue equivalen o P we alo have Z 2 Z = a.e. Similaly, one how ha Y 1 an U a well a Z 1 an Ž coincie o Y 1 = U, Y 2 = V, Z 1 = Ž an Z2 = Z a.e. Now, emembe U = u,x 1 x 1,X 2. Togehe wih u 1,X 1,X 2 = Y 1 an Y 1 = U hi yiel u 1, = u, an, heefoe, u 1 = u a.e. on I x 1 x max. M 1 Similaly, we ge u 2 = u. Now, noe ha u 1 = u i coninuou. Thi make x 2 x 1 Lemma 36 applicable, o Z = Z = U = Y 1 a.e. 14 Theeby Y 1, Y 2 aify he following ynamic: Y 1 = g X 1 T Y 2 = δ X 2 T Z 1 W Z 2 W which implie uing he chain ule of Lemma 5: 1 x iy =g X 1 an T 2 2 x iy = δ X 2 x ix1 x iz1 W T 1 x iy Y 2 +Y 1 T 2 2Y 1 Y 2 Z 1, 15 2Y 1 Y 2 Z 2, [,T, 16 2 x iy x iz2 W x ix2 T 1 x iy Y 2 +Y 1 2 x iy Z 1 +Y 1 Y 2 Z 2 +Y 1 Y 2 x iz1, x iz2, 26

28 fo i = 1,2. Le u ecall ome aemen abou he fowa poce obaine in he poof of Lemma 32: an x 2X2 >, x 2X2 x 1X2 = x 1X1 = 1, 1 = 1 2Y 1 2Y 1 Z 2 Z 1 +Z 2 x 2X1 =, a.e., x 2X2 x 1X2 Uing he chain ule of Lemma 5 an he ecoupling coniion, we have Now, efine i x 1Y = x 1ui,X 1,X 2 i x 2Y = x 2ui,X 1,X 2 Y 12 Y 22 Y 11 Y 21 := x 2u1,X 1,X 2 = x := x 2u2,X 1,X 2 = := x 1u1,X 1,X 2 = x := x 1u2,X 1,X 2 = x + x 2ui,X 1 x 2X2, i = 1,2. 1 2Y 2 x 2Y 1, ,X 2 x 2X2 x 2X2 x 1X2, 1, 19 1, 1 1Y Y 12 x 1X2 2 1Y Y 22 x 1X2., 2 We can apply he Iô fomula o euce ynamic of Y 12 an Y 11 fom ynamic of x 2 Y 1, Le u efine Z 12 := Y 12 = 1, X 2 Y 1 an X 2 : x 2 x 1 x 1 Z 1 1, x 2 X 2 x 2 o we can wie uing 19 +Y 1 Y 2 Z 12 W { 2 x 2Z1 x 2X2 1 x 2Y Y 2 +Y Uing he efiniion of Y 12, Y 22, Z 12 we can implify hi o Y 12 = Z 12 W 2 x 2Y Z 1 1 x 2Y Y 1 Z 2 x 2X2 1 } 2 Y 12 Y 2 +Y 1 Y 22 Z 1 +Y 1 Y 2 Z 12 +Y 12 Y 1 Z

29 Le u now efine Z 11 := Z 1 x 1 Z 12 X 2 x 1, o we can wie uing 2 Y 11 =g X 1 T Z 11 { T 2 2 Y 12 2 Y 1 W 1 x 1Y Y 2 +Y 1 Y 12 Y 2 +Y 1 Y 22 Z 1 +Z 2 2 x 1Y Z 1 +Y 1 Y 2 Z 1 +Y 1 Y 2 Z 12 +Y 12 Y 1 Z 2 }. x 1X2 The wo make em above can be effecively mege ino one uing 2: Y 11 =g X 1 T 2 { 2 Z 11 W Y 11 Y 2 +Y 1 2 x 1Y Z 1 +Y 1 Y 2 Y 1 Y 22 Z 1 +Y 1 Y 2 Z 12 +Y 12 Y 1 Z 2 }. Y 12 2 Y 1 Z 1 +Z 2 x 1X2 x 1Z1 x 1X2 x 1Z1 x 1X2 Similaly, he fou make em can be mege ino only wo uing he ucue of Y 21 an Z 11.. Y 11 =g X 1 T 2 { 2 Z 11 W Y 11 Y 2 +Y 1 Y 21 Z 1 +Y 1 Y 2 Z 11 Y 12 Y 1 Z 2 x 1X2 Y 12 2 Y 1 Z 1 + Z 2 whee he wo make em effecively cancel each ohe ou: Y 11 =g X 1 T 2 Z 11 W } x 1X2, Y 11 Y 2 +Y 1 Y 21 Z 1 +Y 1 Y 2 Z 11 +Y 12 Y 1 Z 1. Analogouly o Y 12 we can euce ynamic of Y 22 : Y 22 = δ X 2 T Z 22 W 2 Y 12 Y 2 +Y 1 Y 22 Z 2 +Y 1 Y 2 Z 22 +Y 22 Y 1 Z 2. 28

30 Fom hee we can, analogouly o Y 11, euce ynamic of Y 21 : Y 21 = Z 21 W 2 Y 11 Y 2 +Y 1 Y 21 Z 2 +Y 1 Y 2 Z 21 +Y 22 Y 1 Z 1. An o we have finally obaine he complee ynamic of he 4-imenional poce Y ij, i,j = 1,2, which ae clealy linea in i. Fuhemoe, emembe: Y 1,Y 2 aeunifomlybouneinepenenlyof,xueoheecouplingconiion, u i = x i u, i = 1,2 an Lemma 32, Z 1, Z 2 ae BMOP pocee wih unifomly boune BMOP-nom inepenenly of,x ue o 15, 16 an Theoem 49, Y ij, i,j = 1,2 ae boune accoing o hei efiniion wih a boun which may epen on,x a hi poin, Z ij, i,j = 1,2 ae in BMOP accoing o Theoem 49, Y ij T i,j=1,2 i unifomly boune by g + δ <. Theefoe, Lemma 48 i applicable an Y ij i,j=1,2 i unifomly boune, inepenenly of,x. In paicula, Y ij = u i,x, i,j = 1,2 can be conolle inepenenly of x j Imax, M x R 2, while u,x, j = 1,2 ha he ame popey a we aleay know. x j Thi how Imax M = [,T uing Lemma 31. Lemma 39. Aume ha g, δ, g, δ ae Lipchiz coninuou. Le u i be he unique i=,1,2 weakly egula Makovian ecoupling fiel o he poblem 13 conuce in Theoem 38. Aume ha u i, i =,1,2, ha a veion which i coninuou in he fi wo componen x 1,x 1 on [,T R 2 fo ome [,T. Then fo any iniial coniion X 1,X 2 = x 1,x 2 = x R 2 he aociae pocee Z i, i =,1,2, on [,T aify Z i ω = x 1ui,X 1 ω,x 2 ω, i =,1,2, fo almo all,ω [,T Ω. Fuhemoe, in hi cae he pocee x 1X2, x 2X2 an 1 x 2X2 on [,T, can be boune unifomly, i.e. inepenenly of, x. Poof. The fi pa of he poof wok analogouly o he poof of Lemma 36. So we keep ou agumen ho. Fo i =,1,2 we conie 1 i E[Y h +h W +h W F 29

31 fo mall h >. A in he poof of Lemma 36, we ue Iô fomula applie o 13 o obain an alo Y i +h W +h W = Y +h W +h W = which lea o an 1 i E[Y h +h +h + +h Y i W + W W Y W + +h W W Z i W 2Z Y 2 Z i + +h 1 E[Y h +h W +h W F = 1 [ +h h E +h W +h W F = 1 h E [ +h W W Z W + F Z +h +h Z i, Z, Z fo h, Z i 1+W W 2Z Y 2 F Z i a h fo i = 1,2. The agumen ae vali fo almo all [,T. On he ohe han we can ue he ecoupling coniion o ewie Y i +h W +h W =u i +h,x 1 +h,x2 W +h W + u i +h,x 1 +h,x2 +h u i +h,x 1 +h,x2 W +h W. Le u eal epaaely wih he wo umman. Fo he fi one ecall ha X 1 an X 2 ae F -meauable, X 1 +h = X1 +W +h W, W +h W i inepenen of F, u i eeminiic, i.e. i aume o be a funcion of,x 1,x 2 [,T R 2. A combinaion of hee popeie lea o 1 [ lim h h E u i +h,x 1 +h,x2 W +h W F = x 1ui,X 1,X 2, if u i i coninuou in he fi wo componen on [,T R 2, whee we ue ha x 1 i boune. Fo he econ umman ecall ha x 1 u i u i i alo Lipchiz coninuou in he la componen wih ome Lipchiz conan L, X 2 +h = X2 + +h 2. Z 3

7 Wave Equation in Higher Dimensions

7 Wave Equation in Higher Dimensions 7 Wave Equaion in Highe Dimensions We now conside he iniial-value poblem fo he wave equaion in n dimensions, u c u x R n u(x, φ(x u (x, ψ(x whee u n i u x i x i. (7. 7. Mehod of Spheical Means Ref: Evans,

More information

Variance and Covariance Processes

Variance and Covariance Processes Vaiance and Covaiance Pocesses Pakash Balachandan Depamen of Mahemaics Duke Univesiy May 26, 2008 These noes ae based on Due s Sochasic Calculus, Revuz and Yo s Coninuous Maingales and Bownian Moion, Kaazas

More information

Thi aicle wa oiginally publihed in a jounal publihed by levie, and he aached copy i povided by levie fo he auho benefi and fo he benefi of he auho iniuion, fo non-commecial eeach and educaional ue including

More information

Online publication date: 01 June 2010 PLEASE SCROLL DOWN FOR ARTICLE

Online publication date: 01 June 2010 PLEASE SCROLL DOWN FOR ARTICLE Thi aicle wa downloaded by: [Yao, Song] On: 7 Januay 211 Acce deail: Acce Deail: [ubcipion numbe 93195159] Publihe Taylo & Fanci Infoma Ld Regieed in England and Wale Regieed Numbe: 172954 Regieed office:

More information

ONTHEPATHWISEUNIQUENESSOF STOCHASTIC PARTIAL DIFFERENTIAL EQUATIONS WITH NON-LIPSCHITZ COEFFICIENTS

ONTHEPATHWISEUNIQUENESSOF STOCHASTIC PARTIAL DIFFERENTIAL EQUATIONS WITH NON-LIPSCHITZ COEFFICIENTS OCKY MOUNTAIN JOUNAL OF MATHEMATICS Volume 43, Numbe 5, 213 ONTHEPATHWISEUNIQUENESSOF STOCHASTIC PATIAL DIFFEENTIAL EQUATIONS WITH NON-LIPSCHITZ COEFFICIENTS DEFEI ZHANG AND PING HE ABSTACT. In hi pape,

More information

Navier Stokes equations and forward backward SDEs on the group of diffeomorphisms of a torus

Navier Stokes equations and forward backward SDEs on the group of diffeomorphisms of a torus Sochaic Pocee and hei Applicaion 9 (9) 434 46 www.elevie.com/locae/pa Navie Soke equaion and fowad backwad SDE on he goup of diffeomophim of a ou Ana Bela Cuzeio a,b,, Evelina Shamaova b,c a Dep. de Maemáica,

More information

General Non-Arbitrage Model. I. Partial Differential Equation for Pricing A. Traded Underlying Security

General Non-Arbitrage Model. I. Partial Differential Equation for Pricing A. Traded Underlying Security 1 Geneal Non-Abiage Model I. Paial Diffeenial Equaion fo Picing A. aded Undelying Secuiy 1. Dynamics of he Asse Given by: a. ds = µ (S, )d + σ (S, )dz b. he asse can be eihe a sock, o a cuency, an index,

More information

Über die Fortpanzung ebener Luftwellen von endlicher Schwingungsweite The Propagation of Planar Air Waves of Finite Amplitude

Über die Fortpanzung ebener Luftwellen von endlicher Schwingungsweite The Propagation of Planar Air Waves of Finite Amplitude Übe ie Fopanzung ebene Lufwellen von enliche Schwingungweie The Popagaion of Plana Ai Wave of Finie Ampliue Benha Riemann Abhanlungen e Geellchaf e Wienchafen zu Göingen 8 pp 43-65 860 Even hough he iffeenial

More information

ENGI 4430 Advanced Calculus for Engineering Faculty of Engineering and Applied Science Problem Set 9 Solutions [Theorems of Gauss and Stokes]

ENGI 4430 Advanced Calculus for Engineering Faculty of Engineering and Applied Science Problem Set 9 Solutions [Theorems of Gauss and Stokes] ENGI 44 Avance alculus fo Engineeing Faculy of Engineeing an Applie cience Poblem e 9 oluions [Theoems of Gauss an okes]. A fla aea A is boune by he iangle whose veices ae he poins P(,, ), Q(,, ) an R(,,

More information

Degree of Approximation of a Class of Function by (C, 1) (E, q) Means of Fourier Series

Degree of Approximation of a Class of Function by (C, 1) (E, q) Means of Fourier Series IAENG Inenaional Jounal of Applied Mahemaic, 4:, IJAM_4 7 Degee of Appoximaion of a Cla of Funcion by C, E, q Mean of Fouie Seie Hae Kihna Nigam and Kuum Shama Abac In hi pape, fo he fi ime, we inoduce

More information

Approximating solutions of backward doubly stochastic differential equations with measurable coefficients using a time discretization scheme

Approximating solutions of backward doubly stochastic differential equations with measurable coefficients using a time discretization scheme Loughboough Univeiy Iniuional Repoioy Appoximaing oluion of backwad doubly ochaic diffeenial equaion wih meauable coefficien uing a ime diceizaion cheme Thi iem wa ubmied o Loughboough Univeiy' Iniuional

More information

A note on characterization related to distributional properties of random translation, contraction and dilation of generalized order statistics

A note on characterization related to distributional properties of random translation, contraction and dilation of generalized order statistics PobSa Foum, Volume 6, July 213, Pages 35 41 ISSN 974-3235 PobSa Foum is an e-jounal. Fo eails please visi www.pobsa.og.in A noe on chaaceizaion elae o isibuional popeies of anom anslaion, conacion an ilaion

More information

Penalization method for a nonlinear Neumann PDE via weak solutions of reflected SDEs

Penalization method for a nonlinear Neumann PDE via weak solutions of reflected SDEs Penalizaion mehod fo a nonlinea Neumann PDE via weak oluion of efleced SDE Khaled Bahlali a,,1, Lucian Maiciuc b,2, Adian Zălinecu b,2, a Univeié de Toulon, IMATH, EA 2134, 83957 La Gade Cedex, Fance.

More information

Time-Space Model of Business Fluctuations

Time-Space Model of Business Fluctuations Time-Sace Moel of Business Flucuaions Aleei Kouglov*, Mahemaical Cene 9 Cown Hill Place, Suie 3, Eobicoke, Onaio M8Y 4C5, Canaa Email: Aleei.Kouglov@SiconVieo.com * This aicle eesens he esonal view of

More information

Stochastic representations of derivatives of solutions of one dimensional parabolic variational inequalities with Neumann boundary conditions

Stochastic representations of derivatives of solutions of one dimensional parabolic variational inequalities with Neumann boundary conditions Sochaic epeenaion of eivaive of oluion of one imenional paabolic vaiaional inequaliie wih Neumann bounay coniion Mieille Boy, Mamaou Cié, Deni alay o cie hi veion: Mieille Boy, Mamaou Cié, Deni alay. Sochaic

More information

Selling at the ultimate maximum in a regime-switching model

Selling at the ultimate maximum in a regime-switching model Selling a he ulimae maximum in a egime-wiching model axiv:158.677v2 mah.pr 24 Jun 216 Yue Liu Nicola Pivaul School of Phyical and Mahemaical Science Diviion of Mahemaical Science Nanyang Technological

More information

Lecture 26: Leapers and Creepers

Lecture 26: Leapers and Creepers Lecue 6: Leape and Ceepe Scibe: Geain Jone (and Main Z. Bazan) Depamen of Economic, MIT May, 5 Inoducion Thi lecue conide he analyi of he non-epaable CTRW in which he diibuion of ep ize and ime beween

More information

Chapter Three Systems of Linear Differential Equations

Chapter Three Systems of Linear Differential Equations Chaper Three Sysems of Linear Differenial Equaions In his chaper we are going o consier sysems of firs orer orinary ifferenial equaions. These are sysems of he form x a x a x a n x n x a x a x a n x n

More information

, the. L and the L. x x. max. i n. It is easy to show that these two norms satisfy the following relation: x x n x = (17.3) max

, the. L and the L. x x. max. i n. It is easy to show that these two norms satisfy the following relation: x x n x = (17.3) max ecure 8 7. Sabiliy Analyi For an n dimenional vecor R n, he and he vecor norm are defined a: = T = i n i (7.) I i eay o how ha hee wo norm aify he following relaion: n (7.) If a vecor i ime-dependen, hen

More information

Selling at the ultimate maximum in a regime-switching model

Selling at the ultimate maximum in a regime-switching model Selling a he ulimae maximum in a egime-wiching model Yue Liu School of Finance and Economic Jiangu Univeiy Zhenjiang 21213 P.R. China Nicola Pivaul School of Phyical and Mahemaical Science Nanyang Technological

More information

Support Vector Machines

Support Vector Machines Suppo Veco Machine CSL 3 ARIFICIAL INELLIGENCE SPRING 4 Suppo Veco Machine O, Kenel Machine Diciminan-baed mehod olean cla boundaie Suppo veco coni of eample cloe o bounday Kenel compue imilaiy beeen eample

More information

Stochastic control for a class of nonlinear kernels and applications *

Stochastic control for a class of nonlinear kernels and applications * Sochaic conol fo a cla of nonlinea kenel and applicaion * Dylan oamaï, Xiaolu Tan, Chao Zhou To cie hi veion: Dylan oamaï, Xiaolu Tan, Chao Zhou. applicaion *. 15. Sochaic conol fo a cla

More information

EECE 301 Signals & Systems Prof. Mark Fowler

EECE 301 Signals & Systems Prof. Mark Fowler EECE 31 Signal & Syem Prof. Mark Fowler Noe Se #27 C-T Syem: Laplace Tranform Power Tool for yem analyi Reading Aignmen: Secion 6.1 6.3 of Kamen and Heck 1/18 Coure Flow Diagram The arrow here how concepual

More information

CHALMERS GÖTEBORG UNIVERSITY

CHALMERS GÖTEBORG UNIVERSITY CHALMERS GÖEBORG UNIVERSIY MASER S HESIS On Weak Diffeeniabiliy of Backwad SDE and Co Hedging of Inuance Deivaive ADAM ANDERSSON Depamen of Mahemaical Saiic CHALMERS UNIVERSIY OF ECHNOLOGY GÖEBORG UNIVERSIY

More information

Laplace Transform. Inverse Laplace Transform. e st f(t)dt. (2)

Laplace Transform. Inverse Laplace Transform. e st f(t)dt. (2) Laplace Tranform Maoud Malek The Laplace ranform i an inegral ranform named in honor of mahemaician and aronomer Pierre-Simon Laplace, who ued he ranform in hi work on probabiliy heory. I i a powerful

More information

PHYS GENERAL RELATIVITY AND COSMOLOGY PROBLEM SET 7 - SOLUTIONS

PHYS GENERAL RELATIVITY AND COSMOLOGY PROBLEM SET 7 - SOLUTIONS PHYS 54 - GENERAL RELATIVITY AND COSMOLOGY - 07 - PROBLEM SET 7 - SOLUTIONS TA: Jeome Quinin Mach, 07 Noe ha houghou hee oluion, we wok in uni whee c, and we chooe he meic ignaue (,,, ) a ou convenion..

More information

Molecular Evolution and Phylogeny. Based on: Durbin et al Chapter 8

Molecular Evolution and Phylogeny. Based on: Durbin et al Chapter 8 Molecula Evoluion and hylogeny Baed on: Dubin e al Chape 8. hylogeneic Tee umpion banch inenal node leaf Topology T : bifucaing Leave - N Inenal node N+ N- Lengh { i } fo each banch hylogeneic ee Topology

More information

Introduction to SLE Lecture Notes

Introduction to SLE Lecture Notes Inroducion o SLE Lecure Noe May 13, 16 - The goal of hi ecion i o find a ufficien condiion of λ for he hull K o be generaed by a imple cure. I urn ou if λ 1 < 4 hen K i generaed by a imple curve. We will

More information

Sections 3.1 and 3.4 Exponential Functions (Growth and Decay)

Sections 3.1 and 3.4 Exponential Functions (Growth and Decay) Secions 3.1 and 3.4 Eponenial Funcions (Gowh and Decay) Chape 3. Secions 1 and 4 Page 1 of 5 Wha Would You Rahe Have... $1million, o double you money evey day fo 31 days saing wih 1cen? Day Cens Day Cens

More information

Representing Knowledge. CS 188: Artificial Intelligence Fall Properties of BNs. Independence? Reachability (the Bayes Ball) Example

Representing Knowledge. CS 188: Artificial Intelligence Fall Properties of BNs. Independence? Reachability (the Bayes Ball) Example C 188: Aificial Inelligence Fall 2007 epesening Knowledge ecue 17: ayes Nes III 10/25/2007 an Klein UC ekeley Popeies of Ns Independence? ayes nes: pecify complex join disibuions using simple local condiional

More information

arxiv: v2 [math.st] 27 Jan 2016

arxiv: v2 [math.st] 27 Jan 2016 STATISTICAL IFERECE VERSUS MEA FIELD LIMIT FOR HAWKES PROCESSES axiv:157.2887v2 mah.st] 27 Jan 216 SYLVAI DELATTRE AD ICOLAS FOURIER Abac. We conide a populaion of individual, of which we obeve he numbe

More information

Randomized Perfect Bipartite Matching

Randomized Perfect Bipartite Matching Inenive Algorihm Lecure 24 Randomized Perfec Biparie Maching Lecurer: Daniel A. Spielman April 9, 208 24. Inroducion We explain a randomized algorihm by Ahih Goel, Michael Kapralov and Sanjeev Khanna for

More information

Algorithmic Discrete Mathematics 6. Exercise Sheet

Algorithmic Discrete Mathematics 6. Exercise Sheet Algorihmic Dicree Mahemaic. Exercie Shee Deparmen of Mahemaic SS 0 PD Dr. Ulf Lorenz 7. and 8. Juni 0 Dipl.-Mah. David Meffer Verion of June, 0 Groupwork Exercie G (Heap-Sor) Ue Heap-Sor wih a min-heap

More information

Reinforcement learning

Reinforcement learning Lecue 3 Reinfocemen leaning Milos Hauskech milos@cs.pi.edu 539 Senno Squae Reinfocemen leaning We wan o lean he conol policy: : X A We see examples of x (bu oupus a ae no given) Insead of a we ge a feedback

More information

Deviation probability bounds for fractional martingales and related remarks

Deviation probability bounds for fractional martingales and related remarks Deviaion pobabiliy bounds fo facional maingales and elaed emaks Buno Sausseeau Laboaoie de Mahémaiques de Besançon CNRS, UMR 6623 16 Roue de Gay 253 Besançon cedex, Fance Absac In his pape we pove exponenial

More information

Chapter 9 - The Laplace Transform

Chapter 9 - The Laplace Transform Chaper 9 - The Laplace Tranform Selece Soluion. Skech he pole-zero plo an region of convergence (if i exi) for hee ignal. ω [] () 8 (a) x e u = 8 ROC σ ( ) 3 (b) x e co π u ω [] ( ) () (c) x e u e u ROC

More information

On Control Problem Described by Infinite System of First-Order Differential Equations

On Control Problem Described by Infinite System of First-Order Differential Equations Ausalian Jounal of Basic and Applied Sciences 5(): 736-74 ISS 99-878 On Conol Poblem Descibed by Infinie Sysem of Fis-Ode Diffeenial Equaions Gafujan Ibagimov and Abbas Badaaya J'afau Insiue fo Mahemaical

More information

8.5 Circles and Lengths of Segments

8.5 Circles and Lengths of Segments LenghofSegmen20052006.nb 1 8.5 Cicle and Lengh of Segmen In hi ecion we will how (and in ome cae pove) ha lengh of chod, ecan, and angen ae elaed in ome nal way. We will look a hee heoem ha ae hee elaionhip

More information

Then the number of elements of S of weight n is exactly the number of compositions of n into k parts.

Then the number of elements of S of weight n is exactly the number of compositions of n into k parts. Geneating Function In a geneal combinatoial poblem, we have a univee S of object, and we want to count the numbe of object with a cetain popety. Fo example, if S i the et of all gaph, we might want to

More information

Main Reference: Sections in CLRS.

Main Reference: Sections in CLRS. Maximum Flow Reied 09/09/200 Main Reference: Secion 26.-26. in CLRS. Inroducion Definiion Muli-Source Muli-Sink The Ford-Fulkeron Mehod Reidual Nework Augmening Pah The Max-Flow Min-Cu Theorem The Edmond-Karp

More information

CSC 364S Notes University of Toronto, Spring, The networks we will consider are directed graphs, where each edge has associated with it

CSC 364S Notes University of Toronto, Spring, The networks we will consider are directed graphs, where each edge has associated with it CSC 36S Noe Univeriy of Torono, Spring, 2003 Flow Algorihm The nework we will conider are direced graph, where each edge ha aociaed wih i a nonnegaive capaciy. The inuiion i ha if edge (u; v) ha capaciy

More information

Graphs III - Network Flow

Graphs III - Network Flow Graph III - Nework Flow Flow nework eup graph G=(V,E) edge capaciy w(u,v) 0 - if edge doe no exi, hen w(u,v)=0 pecial verice: ource verex ; ink verex - no edge ino and no edge ou of Aume every verex v

More information

Network Flows: Introduction & Maximum Flow

Network Flows: Introduction & Maximum Flow CSC 373 - lgorihm Deign, nalyi, and Complexiy Summer 2016 Lalla Mouaadid Nework Flow: Inroducion & Maximum Flow We now urn our aenion o anoher powerful algorihmic echnique: Local Search. In a local earch

More information

Lecture 22 Electromagnetic Waves

Lecture 22 Electromagnetic Waves Lecue Elecomagneic Waves Pogam: 1. Enegy caied by he wave (Poyning veco).. Maxwell s equaions and Bounday condiions a inefaces. 3. Maeials boundaies: eflecion and efacion. Snell s Law. Quesions you should

More information

The distribution of the interval of the Cox process with shot noise intensity for insurance claims and its moments

The distribution of the interval of the Cox process with shot noise intensity for insurance claims and its moments The diibuion of he ineval of he Cox poce wih ho noie ineniy fo inuance claim and i momen Angelo Daio, Ji-Wook Jang Depamen of Saiic, London School of Economic and Poliical Science, Houghon See, London

More information

Explicit form of global solution to stochastic logistic differential equation and related topics

Explicit form of global solution to stochastic logistic differential equation and related topics SAISICS, OPIMIZAION AND INFOMAION COMPUING Sa., Opim. Inf. Compu., Vol. 5, March 17, pp 58 64. Publihed online in Inernaional Academic Pre (www.iapre.org) Explici form of global oluion o ochaic logiic

More information

Computer Propagation Analysis Tools

Computer Propagation Analysis Tools Compue Popagaion Analysis Tools. Compue Popagaion Analysis Tools Inoducion By now you ae pobably geing he idea ha pedicing eceived signal sengh is a eally impoan as in he design of a wieless communicaion

More information

18.03SC Unit 3 Practice Exam and Solutions

18.03SC Unit 3 Practice Exam and Solutions Sudy Guide on Sep, Dela, Convoluion, Laplace You can hink of he ep funcion u() a any nice mooh funcion which i for < a and for > a, where a i a poiive number which i much maller han any ime cale we care

More information

Discussion Session 2 Constant Acceleration/Relative Motion Week 03

Discussion Session 2 Constant Acceleration/Relative Motion Week 03 PHYS 100 Dicuion Seion Conan Acceleraion/Relaive Moion Week 03 The Plan Today you will work wih your group explore he idea of reference frame (i.e. relaive moion) and moion wih conan acceleraion. You ll

More information

MEEN 617 Handout #11 MODAL ANALYSIS OF MDOF Systems with VISCOUS DAMPING

MEEN 617 Handout #11 MODAL ANALYSIS OF MDOF Systems with VISCOUS DAMPING MEEN 67 Handou # MODAL ANALYSIS OF MDOF Sysems wih VISCOS DAMPING ^ Symmeic Moion of a n-dof linea sysem is descibed by he second ode diffeenial equaions M+C+K=F whee () and F () ae n ows vecos of displacemens

More information

Control Volume Derivation

Control Volume Derivation School of eospace Engineeing Conol Volume -1 Copyigh 1 by Jey M. Seizman. ll ighs esee. Conol Volume Deiaion How o cone ou elaionships fo a close sysem (conol mass) o an open sysem (conol olume) Fo mass

More information

UT Austin, ECE Department VLSI Design 5. CMOS Gate Characteristics

UT Austin, ECE Department VLSI Design 5. CMOS Gate Characteristics La moule: CMOS Tranior heory Thi moule: DC epone Logic Level an Noie Margin Tranien epone Delay Eimaion Tranior ehavior 1) If he wih of a ranior increae, he curren will ) If he lengh of a ranior increae,

More information

6.8 Laplace Transform: General Formulas

6.8 Laplace Transform: General Formulas 48 HAP. 6 Laplace Tranform 6.8 Laplace Tranform: General Formula Formula Name, ommen Sec. F() l{ f ()} e f () d f () l {F()} Definiion of Tranform Invere Tranform 6. l{af () bg()} al{f ()} bl{g()} Lineariy

More information

Lecture-V Stochastic Processes and the Basic Term-Structure Equation 1 Stochastic Processes Any variable whose value changes over time in an uncertain

Lecture-V Stochastic Processes and the Basic Term-Structure Equation 1 Stochastic Processes Any variable whose value changes over time in an uncertain Lecue-V Sochasic Pocesses and he Basic Tem-Sucue Equaion 1 Sochasic Pocesses Any vaiable whose value changes ove ime in an unceain way is called a Sochasic Pocess. Sochasic Pocesses can be classied as

More information

D.I. Survival models and copulas

D.I. Survival models and copulas D- D. SURVIVAL COPULA D.I. Survival moels an copulas Definiions, relaionships wih mulivariae survival isribuion funcions an relaionships beween copulas an survival copulas. D.II. Fraily moels Use of a

More information

Sharif University of Technology - CEDRA By: Professor Ali Meghdari

Sharif University of Technology - CEDRA By: Professor Ali Meghdari Shaif Univesiy of echnology - CEDRA By: Pofesso Ali Meghai Pupose: o exen he Enegy appoach in eiving euaions of oion i.e. Lagange s Meho fo Mechanical Syses. opics: Genealize Cooinaes Lagangian Euaion

More information

Generalized Orlicz Spaces and Wasserstein Distances for Convex-Concave Scale Functions

Generalized Orlicz Spaces and Wasserstein Distances for Convex-Concave Scale Functions Generalized Orlicz Space and Waerein Diance for Convex-Concave Scale Funcion Karl-Theodor Surm Abrac Given a ricly increaing, coninuou funcion ϑ : R + R +, baed on he co funcional ϑ (d(x, y dq(x, y, we

More information

Design of Controller for Robot Position Control

Design of Controller for Robot Position Control eign of Conroller for Robo oiion Conrol Two imporan goal of conrol: 1. Reference inpu racking: The oupu mu follow he reference inpu rajecory a quickly a poible. Se-poin racking: Tracking when he reference

More information

To become more mathematically correct, Circuit equations are Algebraic Differential equations. from KVL, KCL from the constitutive relationship

To become more mathematically correct, Circuit equations are Algebraic Differential equations. from KVL, KCL from the constitutive relationship Laplace Tranform (Lin & DeCarlo: Ch 3) ENSC30 Elecric Circui II The Laplace ranform i an inegral ranformaion. I ranform: f ( ) F( ) ime variable complex variable From Euler > Lagrange > Laplace. Hence,

More information

The shortest path between two truths in the real domain passes through the complex domain. J. Hadamard

The shortest path between two truths in the real domain passes through the complex domain. J. Hadamard Complex Analysis R.G. Halbud R.Halbud@ucl.ac.uk Depamen of Mahemaics Univesiy College London 202 The shoes pah beween wo uhs in he eal domain passes hough he complex domain. J. Hadamad Chape The fis fundamenal

More information

1 Motivation and Basic Definitions

1 Motivation and Basic Definitions CSCE : Deign and Analyi of Algorihm Noe on Max Flow Fall 20 (Baed on he preenaion in Chaper 26 of Inroducion o Algorihm, 3rd Ed. by Cormen, Leieron, Rive and Sein.) Moivaion and Baic Definiion Conider

More information

The Production of Well-Being: Conventional Goods, Relational Goods and Status Goods

The Production of Well-Being: Conventional Goods, Relational Goods and Status Goods The Poducion of Well-Bein: Convenional Good, Relaional Good and Sau Good Aloy Pinz Iniue of Public Economic II Univeiy of Müne, Gemany New Diecion in Welfae II, OECD Pai July 06 08, 2011 Conen 1. Inoducion

More information

Fall 2004/05 Solutions to Assignment 5: The Stationary Phase Method Provided by Mustafa Sabri Kilic. I(x) = e ixt e it5 /5 dt (1) Z J(λ) =

Fall 2004/05 Solutions to Assignment 5: The Stationary Phase Method Provided by Mustafa Sabri Kilic. I(x) = e ixt e it5 /5 dt (1) Z J(λ) = 8.35 Fall 24/5 Solution to Aignment 5: The Stationay Phae Method Povided by Mutafa Sabi Kilic. Find the leading tem fo each of the integal below fo λ >>. (a) R eiλt3 dt (b) R e iλt2 dt (c) R eiλ co t dt

More information

The Residual Graph. 11 Augmenting Path Algorithms. Augmenting Path Algorithm. Augmenting Path Algorithm

The Residual Graph. 11 Augmenting Path Algorithms. Augmenting Path Algorithm. Augmenting Path Algorithm Augmening Pah Algorihm Greedy-algorihm: ar wih f (e) = everywhere find an - pah wih f (e) < c(e) on every edge augmen flow along he pah repea a long a poible The Reidual Graph From he graph G = (V, E,

More information

SPHERICAL WINDS SPHERICAL ACCRETION

SPHERICAL WINDS SPHERICAL ACCRETION SPHERICAL WINDS SPHERICAL ACCRETION Spheical wins. Many sas ae known o loose mass. The sola win caies away abou 10 14 M y 1 of vey ho plasma. This ae is insignifican. In fac, sola aiaion caies away 4 10

More information

The Residual Graph. 12 Augmenting Path Algorithms. Augmenting Path Algorithm. Augmenting Path Algorithm

The Residual Graph. 12 Augmenting Path Algorithms. Augmenting Path Algorithm. Augmenting Path Algorithm Augmening Pah Algorihm Greedy-algorihm: ar wih f (e) = everywhere find an - pah wih f (e) < c(e) on every edge augmen flow along he pah repea a long a poible The Reidual Graph From he graph G = (V, E,

More information

Chapter 2 The Derivative Applied Calculus 107. We ll need a rule for finding the derivative of a product so we don t have to multiply everything out.

Chapter 2 The Derivative Applied Calculus 107. We ll need a rule for finding the derivative of a product so we don t have to multiply everything out. Chaper The Derivaive Applie Calculus 107 Secion 4: Prouc an Quoien Rules The basic rules will le us ackle simple funcions. Bu wha happens if we nee he erivaive of a combinaion of hese funcions? Eample

More information

The Fundamental Theorems of Calculus

The Fundamental Theorems of Calculus FunamenalTheorems.nb 1 The Funamenal Theorems of Calculus You have now been inrouce o he wo main branches of calculus: ifferenial calculus (which we inrouce wih he angen line problem) an inegral calculus

More information

Flow networks. Flow Networks. A flow on a network. Flow networks. The maximum-flow problem. Introduction to Algorithms, Lecture 22 December 5, 2001

Flow networks. Flow Networks. A flow on a network. Flow networks. The maximum-flow problem. Introduction to Algorithms, Lecture 22 December 5, 2001 CS 545 Flow Nework lon Efra Slide courey of Charle Leieron wih mall change by Carola Wenk Flow nework Definiion. flow nework i a direced graph G = (V, E) wih wo diinguihed verice: a ource and a ink. Each

More information

Exponential and Logarithmic Equations and Properties of Logarithms. Properties. Properties. log. Exponential. Logarithmic.

Exponential and Logarithmic Equations and Properties of Logarithms. Properties. Properties. log. Exponential. Logarithmic. Eponenial and Logaihmic Equaions and Popeies of Logaihms Popeies Eponenial a a s = a +s a /a s = a -s (a ) s = a s a b = (ab) Logaihmic log s = log + logs log/s = log - logs log s = s log log a b = loga

More information

Optimal Decentralized State-Feedback Control with Sparsity and Delays

Optimal Decentralized State-Feedback Control with Sparsity and Delays Opimal Decenalized Sae-Feedback Conol wih Spaiy and Delay ndew Lampeki Lauen Lead uomaica, vol. 58, pp. 43 5, ug. 5 bac Thi wok peen he oluion o a cla of decenalized linea quadaic ae-feedback conol poblem,

More information

Stability in Distribution for Backward Uncertain Differential Equation

Stability in Distribution for Backward Uncertain Differential Equation Sabiliy in Diribuion for Backward Uncerain Differenial Equaion Yuhong Sheng 1, Dan A. Ralecu 2 1. College of Mahemaical and Syem Science, Xinjiang Univeriy, Urumqi 8346, China heng-yh12@mail.inghua.edu.cn

More information

Coupled Mass Transport and Reaction in LPCVD Reactors

Coupled Mass Transport and Reaction in LPCVD Reactors ople Ma Tanpo an eaion in LPV eao ile A in B e.g., SiH 4 in H Sepaae eao ino o egion, inaafe & annla b - oniniy Eqn: : onveion-iffion iffion-eaion Eqn Ampion! ile peie i in majo aie ga e.g., H isih 4!

More information

The Clique Density Theorem.

The Clique Density Theorem. Annal of Mahemaic 00 (XXXX), 28 doi: 0.4007/annal.XXXX.00.0.0000 The Clique Deniy Theoem. By Chiian Reihe Abac Conide he following queion: Fix an inege 3 a well a a eal paamee γ [0, /2) and le n be a lage

More information

Problem Set If all directed edges in a network have distinct capacities, then there is a unique maximum flow.

Problem Set If all directed edges in a network have distinct capacities, then there is a unique maximum flow. CSE 202: Deign and Analyi of Algorihm Winer 2013 Problem Se 3 Inrucor: Kamalika Chaudhuri Due on: Tue. Feb 26, 2013 Inrucion For your proof, you may ue any lower bound, algorihm or daa rucure from he ex

More information

u(t) Figure 1. Open loop control system

u(t) Figure 1. Open loop control system Open loop conrol v cloed loop feedbac conrol The nex wo figure preen he rucure of open loop and feedbac conrol yem Figure how an open loop conrol yem whoe funcion i o caue he oupu y o follow he reference

More information

Exponential Sawtooth

Exponential Sawtooth ECPE 36 HOMEWORK 3: PROPERTIES OF THE FOURIER TRANSFORM SOLUTION. Exponenial Sawooh: The eaie way o do hi problem i o look a he Fourier ranform of a ingle exponenial funcion, () = exp( )u(). From he able

More information

CHAPTER 7: SECOND-ORDER CIRCUITS

CHAPTER 7: SECOND-ORDER CIRCUITS EEE5: CI RCUI T THEORY CHAPTER 7: SECOND-ORDER CIRCUITS 7. Inroducion Thi chaper conider circui wih wo orage elemen. Known a econd-order circui becaue heir repone are decribed by differenial equaion ha

More information

Consider a Binary antipodal system which produces data of δ (t)

Consider a Binary antipodal system which produces data of δ (t) Modulaion Polem PSK: (inay Phae-hi keying) Conide a inay anipodal yem whih podue daa o δ ( o + δ ( o inay and epeively. Thi daa i paed o pule haping ile and he oupu o he pule haping ile i muliplied y o(

More information

Introduction to Congestion Games

Introduction to Congestion Games Algorihmic Game Theory, Summer 2017 Inroducion o Congeion Game Lecure 1 (5 page) Inrucor: Thoma Keelheim In hi lecure, we ge o know congeion game, which will be our running example for many concep in game

More information

Let. x y. denote a bivariate time series with zero mean.

Let. x y. denote a bivariate time series with zero mean. Linear Filer Le x y : T denoe a bivariae ime erie wih zero mean. Suppoe ha he ime erie {y : T} i conruced a follow: y a x The ime erie {y : T} i aid o be conruced from {x : T} by mean of a Linear Filer.

More information

Research on Simplifying the Motion Equations for Submarine Training Simulator based on Sensitivity Index

Research on Simplifying the Motion Equations for Submarine Training Simulator based on Sensitivity Index Secon nenaional Sympoium on nelligen nfomaion echnology Alicaion Reeach on Simplifying he Moion Euaion fo Submaine aining Simulao bae on Seniiiy nex Zhao Lin,Zhu Yi,Lu Zhihong 3,Zhang Jing 4 College of

More information

Fractional Brownian Bridge Measures and Their Integration by Parts Formula

Fractional Brownian Bridge Measures and Their Integration by Parts Formula Journal of Mahemaical Reearch wih Applicaion Jul., 218, Vol. 38, No. 4, pp. 418 426 DOI:1.377/j.in:295-2651.218.4.9 Hp://jmre.lu.eu.cn Fracional Brownian Brige Meaure an Their Inegraion by Par Formula

More information

ESTIMATES FOR THE DERIVATIVE OF DIFFUSION SEMIGROUPS

ESTIMATES FOR THE DERIVATIVE OF DIFFUSION SEMIGROUPS Elec. Comm. in Probab. 3 (998) 65 74 ELECTRONIC COMMUNICATIONS in PROBABILITY ESTIMATES FOR THE DERIVATIVE OF DIFFUSION SEMIGROUPS L.A. RINCON Deparmen of Mahemaic Univeriy of Wale Swanea Singleon Par

More information

Physics 240: Worksheet 16 Name

Physics 240: Worksheet 16 Name Phyic 4: Workhee 16 Nae Non-unifor circular oion Each of hee proble involve non-unifor circular oion wih a conan α. (1) Obain each of he equaion of oion for non-unifor circular oion under a conan acceleraion,

More information

On the quadratic support of strongly convex functions

On the quadratic support of strongly convex functions Int. J. Nonlinea Anal. Appl. 7 2016 No. 1, 15-20 ISSN: 2008-6822 electonic http://dx.doi.og/10.22075/ijnaa.2015.273 On the quadatic uppot of tongly convex function S. Abbazadeh a,b,, M. Ehaghi Godji a

More information

Single Phase Line Frequency Uncontrolled Rectifiers

Single Phase Line Frequency Uncontrolled Rectifiers Single Phae Line Frequency Unconrolle Recifier Kevin Gaughan 24-Nov-03 Single Phae Unconrolle Recifier 1 Topic Baic operaion an Waveform (nucive Loa) Power Facor Calculaion Supply curren Harmonic an Th

More information

Fractional Ornstein-Uhlenbeck Bridge

Fractional Ornstein-Uhlenbeck Bridge WDS'1 Proceeding of Conribued Paper, Par I, 21 26, 21. ISBN 978-8-7378-139-2 MATFYZPRESS Fracional Ornein-Uhlenbeck Bridge J. Janák Charle Univeriy, Faculy of Mahemaic and Phyic, Prague, Czech Republic.

More information

Risk tolerance and optimal portfolio choice

Risk tolerance and optimal portfolio choice Risk oleance and opimal pofolio choice Maek Musiela BNP Paibas London Copoae and Invesmen Join wok wih T. Zaiphopoulou (UT usin) Invesmens and fowad uiliies Pepin 6 Backwad and fowad dynamic uiliies and

More information

, on the power of the transmitter P t fed to it, and on the distance R between the antenna and the observation point as. r r t

, on the power of the transmitter P t fed to it, and on the distance R between the antenna and the observation point as. r r t Lecue 6: Fiis Tansmission Equaion and Rada Range Equaion (Fiis equaion. Maximum ange of a wieless link. Rada coss secion. Rada equaion. Maximum ange of a ada. 1. Fiis ansmission equaion Fiis ansmission

More information

Algorithms and Data Structures 2011/12 Week 9 Solutions (Tues 15th - Fri 18th Nov)

Algorithms and Data Structures 2011/12 Week 9 Solutions (Tues 15th - Fri 18th Nov) Algorihm and Daa Srucure 2011/ Week Soluion (Tue 15h - Fri 18h No) 1. Queion: e are gien 11/16 / 15/20 8/13 0/ 1/ / 11/1 / / To queion: (a) Find a pair of ube X, Y V uch ha f(x, Y) = f(v X, Y). (b) Find

More information

Notes on cointegration of real interest rates and real exchange rates. ρ (2)

Notes on cointegration of real interest rates and real exchange rates. ρ (2) Noe on coinegraion of real inere rae and real exchange rae Charle ngel, Univeriy of Wiconin Le me ar wih he obervaion ha while he lieraure (mo prominenly Meee and Rogoff (988) and dion and Paul (993))

More information

FI 2201 Electromagnetism

FI 2201 Electromagnetism FI Electomagnetim Aleande A. Ikanda, Ph.D. Phyic of Magnetim and Photonic Reeach Goup ecto Analyi CURILINEAR COORDINAES, DIRAC DELA FUNCION AND HEORY OF ECOR FIELDS Cuvilinea Coodinate Sytem Cateian coodinate:

More information

A STOCHASTIC MODELING FOR THE UNSTABLE FINANCIAL MARKETS

A STOCHASTIC MODELING FOR THE UNSTABLE FINANCIAL MARKETS A STOCHASTIC MODELING FOR THE UNSTABLE FINANCIAL MARKETS Assoc. Pof. Romeo Negea Ph. D Poliehnica Univesiy of Timisoaa Depamen of Mahemaics Timisoaa, Romania Assoc. Pof. Cipian Peda Ph. D Wes Univesiy

More information

CS4445/9544 Analysis of Algorithms II Solution for Assignment 1

CS4445/9544 Analysis of Algorithms II Solution for Assignment 1 Conider he following flow nework CS444/944 Analyi of Algorihm II Soluion for Aignmen (0 mark) In he following nework a minimum cu ha capaciy 0 Eiher prove ha hi aemen i rue, or how ha i i fale Uing he

More information

Algorithm Design and Analysis

Algorithm Design and Analysis Algorihm Deign and Analyi LECTURE 0 Nework Flow Applicaion Biparie maching Edge-dijoin pah Adam Smih 0//0 A. Smih; baed on lide by E. Demaine, C. Leieron, S. Rakhodnikova, K. Wayne La ime: Ford-Fulkeron

More information

156 There are 9 books stacked on a shelf. The thickness of each book is either 1 inch or 2

156 There are 9 books stacked on a shelf. The thickness of each book is either 1 inch or 2 156 Thee ae 9 books sacked on a shelf. The hickness of each book is eihe 1 inch o 2 F inches. The heigh of he sack of 9 books is 14 inches. Which sysem of equaions can be used o deemine x, he numbe of

More information

Mathematische Annalen

Mathematische Annalen Mah. Ann. 39, 33 339 (997) Mahemaiche Annalen c Springer-Verlag 997 Inegraion by par in loop pace Elon P. Hu Deparmen of Mahemaic, Norhweern Univeriy, Evanon, IL 628, USA (e-mail: elon@@mah.nwu.edu) Received:

More information

NECESSARY AND SUFFICIENT CONDITIONS FOR LATENT SEPARABILITY

NECESSARY AND SUFFICIENT CONDITIONS FOR LATENT SEPARABILITY NECESSARY AND SUFFICIENT CONDITIONS FOR LATENT SEPARABILITY Ian Crawford THE INSTITUTE FOR FISCAL STUDIES DEPARTMENT OF ECONOMICS, UCL cemmap working paper CWP02/04 Neceary and Sufficien Condiion for Laen

More information

The sudden release of a large amount of energy E into a background fluid of density

The sudden release of a large amount of energy E into a background fluid of density 10 Poin explosion The sudden elease of a lage amoun of enegy E ino a backgound fluid of densiy ceaes a song explosion, chaaceized by a song shock wave (a blas wave ) emanaing fom he poin whee he enegy

More information