Towards the Fusion of Distributed Binary Decision Tree Classifiers

Size: px
Start display at page:

Download "Towards the Fusion of Distributed Binary Decision Tree Classifiers"

Transcription

1 Towads th Fusion of Distibutd Binay Dcision T Cassifis Qian Zhang EECS Dpatmnt, Link Ha Syacus Univsity, Syacus, NY 344, USA Emai: qizhang@sy.du Abstact Mutip snso fusion and binay dcision t cassifis hav bn usd to succssfuy sov many a wod pobms. Ths topics a usuay studid spaaty. Fusion of binay dcision t cassifis in a mutip snso nvionmnt has civd vy itt attntion. In this pap, w fomuat th pobm, invstigat its scop, outin som issus associatd with dcision t cassifis and mutip snso fusion, and psnt som soution mthodoogis. Th suts a iustatd by mans of an xamp. Ky wods: snso fusion, binay dcision t. I. Intoduction Mutip snso dcision fusion is an impotant pobm with many pactica appications. This pobm has bn studid quit xtnsivy and many significant suts on this topic hav bn obtaind [- 3]. In most studis, on-stag dcision making pocdus a mpoyd at th snso as w as at th fusion cnt. By a on-stag pocdu, w man that a sing tst is mpoyd to distinguish btwn a th hypothss. Such on-stag dcision pocdus may bcom too compx and impactica in situations wh th a many vaiabs and many objct casss (hypothss). Ths typs of pobms ais in aas such as automatic tagt cognition (ATR), tam mdica diagnosis and tmdicin, and agsca suvianc systms. In ths situations, it may b dsiab to us mutistag dcision making pocdus at th snsos and/o at th fusion cnt. Th a sva appoachs fo impmnting mutistag dcision making stuctus. On popua appoach is by mans of a binay dcision t (BDT) [4]. Th basic ida is to tak a compx M-ay hypothsis tsting pobm, bak it into sva simp binay hypothsis tsting pobms that a oganizd in a hiachica t stuctu to mak dcisions gading th M hypothss o objct casss. H, w invstigat th us of BDTs in mutisnso fusion pobms. Th bnfits of using BDTs in mutisnso dcision fusion a mutifod:. It is w known that th dsign of distibutd dtction systms that mpoy on-stag dcision making fo binay hypothsis tsting pobms is Pamod K. Vashny EECS Dpatmnt, Link Ha Syacus Univsity, Syacus, NY 344, USA Emai: vashny@sy.du NP-had. This dsign fo M-ay hypothsis tsting is vn had. BDTs mak a squnc of binay dcisions in a hiachica mann that a asi to dsign, fficint and computationay simp to impmnt with simp dcision gions. Thus, th us of a BDT may mak th dcision making pocdu fasib fo pactica situations that hav tim o pocssing constaints. In addition, communication bandwidth fficincy may b achivd bcaus tansmission of binay dcisions instad of M-ay dcisions wi b quid.. Som of th avaiab snsos may not b capab of distinguishing a th objct casss in a paiwis mann. BDTs povid a famwok fo intgating th capabiity of a th snsos fo mutisnso dcision fusion and fo nhancd systm pfomanc. 3. Dcision making via a BDT has an inhnt fxibiity to dsign tsts at th intna nods. This fxibiity povids th abiity to hand snso dfcts, missing snso obsvations/dcisions, tc., thby nhancing systm obustnss. Aso, th fxibiity may hp in impoving systm pfomanc. Th dsign of a BDT basd mutisnso fusion systm invovs th dsign of th BDT, dsign of th tsts at th intna nods of snso BDTs, dsign of th fusion u, and dsign of th systm topoogy incuding communication stuctu of th mutisnso fusion systm. Goas of this dsign incud nhancd ova systm pfomanc (cognition abiity) and obustnss, using ast possib computation and communication. Som aspcts of this pobm hav bn addssd in th itatu. Dmibas [5] poposd a non-paamtic cntaizd objct cognition schm basd on a BDT. Each snso pocsss its data and xtacts som fatus that a tansmittd to th fusion cnt. Objct cognition is caid out using a BDT gnatd fom a taining st. Dasaathy [6] concntatd on th achitctua aspcts of a systm that fuss binay dcisions into a sing M- ay dcision. Th main goa was to dsign achitctus that satisfy pocssing tim constaints. Zhu t a. [7] aso considd th pobm of M-ay hypothsis tsting using a paa fusion topoogy wh oca dtctos tansmit binay dcisions. Thy focussd on th dsign of dcision us and on systm

2 pfomanc. Th main goa of this pap is to xamin th ova pobm of BDT basd mutisnso dcision fusion pobm, idntify th issus that nd to b addssd futh and popos som soution mthodoogis. In Sction II, w fomuat th pobm. In Sction III, w consid a spcific BDT basd dcision fusion systm. In Sction IV, w giv an xamp to iustat th suts of Sction III. In Sction V, w mak som concuding maks. II. Pobm Ovviw W focus ou attntion on th paa achitctu fo BDT basd mutisnso fusion systms in this pap. Th bock diagam of a BDT basd paa dcision fusion systm is shown in Figu. Th systm consists of K snsos that obsv a common phnomnon in paa. Th goa is to cogniz a givn unknown objct that bongs to th st of objcts {O, O,, O M. Lt X k b th obsvation vcto of th kth snso. Each snso uss a BDT to mak its dcisions. Lt T k dnot th BDT usd by th kth snso and U k th dcision mad by th kth snso. Ths dcisions a tansmittd to th fusion cnt that combins thm to yid U, th goba dcision. Th fusion u Γ (.) is a function that maps oca dcisions u,, u K into u and maks a dcision gading th unknown objct. X Snso S BDT T U U =Γ (U,,U K ) U U k X K Snso S K BDT T K Figu : BDT basd paa dcision fusion systm W assum that th ad is famiia with th notion of t and associatd tminoogy. Fo dtais, th ad may f to [4]. A gna BDT is shown in Figu. X dnots th fatu vcto. U dnots th dcision mad by th BDT at tmina nods. Sinc ach vau u of U cosponds to a uniqu path fom th oot nod to a tmina nod, u can b ncodd as th squnc of binay dcisions mad by a th nods in th cosponding path. At nod t, Φ(t) dnots th st of fatus usd by th BDT, and Γ(t) dnots th dcision u, which is a function that maps th spac of fatus spcifid by Φ(t) into st {,. nod t Φ (t), Γ(t) X oot O a U=u a O b U=u b Figu : A gna BDT Using th basic systm achitctu fo th dcision fusion systm shown in Figu, sva dsign appoachs and mods of opation can b nvisagd. H w catgoiz BDT basd paa dcision fusion systms in fou typs:. Each snso mpoys a BDT fo dcision making. Ths snso BDTs a assumd avaiab and a dsignd indpndnty of ach oth. Snso dcisions a snt to th fusion cnt. Th fusion cnt ith uss a on-stag pocdu o a BDT to dtmin th fina dcision. Th main issu h is th dsign of an optimum fusion pocdu. This pobm is anaogous to th dsign of an optimum fusion u in distibutd dtction systms [9]. This pobm is appicab to th fusion of BDT cassifis that may hav bn dsignd indpndnty.. BDTs at th individua snsos a dsignd jointy by mpoying coupd cost functions. Dcisions a avaiab ocay so that appopiat action can b takn at th snso. Dcision fusion is not mpoyd to combin snso dcisions. This pobm is anaogous to th on considd by Tnny and Sand [] in a distibutd dtction contxt. 3. Dcisions mad by snso BDTs a convyd to th fusion cnt that combins thm to yid th fina dcision. No oth communication is aowd. Snso BDTs and th fusion u a dsignd. This systm is a gnaization of distibutd dtction systms wh ony on-stag dcision pocdus a aowd. 4. In this cas, th systm is th sam as systm 3 xcpt that two-way communication btwn th snsos and th fusion cnt is aowd. Th

3 snsos navigat though thi BDTs und th supvision of th fusion cnt in a tuy coopativ mann. This is a nw systm achitctu that quis cos coodination btwn systm mnts. In th fou typs of systms dscibd abov, som o a of th foowing issus nd to b addssd. Dign of BDTs fo ach snso and/o th fusion cnt. Sction of snso fatus fo snso BDT dsign and dcision-making at intna nods of snso BDTs. Dcision-making us at intna nods of snso BDTs. Dcision-making us at th fusion cnt. Communication potoco usd by th snsos and th fusion cnt fo coodinatd navigation of thi BDTs. Evauation of systm pfomanc - cognition abiity, obustnss tc. Th tatmnt of th abov issus is guidd by many factos, such as snso chaactistics, avaiab pocssing soucs, opating nvionmnt and th natu of th objcts, tc. Du to th novty of ths pobms, much ffot is ndd to dvop a systmatic soution to th ova BDT basd mutisnso fusion pobm. In th foowing w mak som obsvations on th diffncs btwn how th abov issus a tatd in th convntiona sns and und th fomuation of BDT basd dcision fusion. Convntiona BDT dsign mthodoogis can b appid h, but with th goa of optimizing th ova systm pfomanc instad of just th pfomanc of individua snsos. A BDTs shoud b dsignd in such a way that thy coctivy mak th bst cassification of th unknown objct. A sing snso BDT may not b optima whn usd as a standaon cassifi. Snso fatus ought to b sctd in such a way that objcts a bst spaatd acoss th snso suit fo nhancd pfomanc of th ova systm. Th sctd fatus may not b th bst fo individua snsos. Th dcision us at intna nods of snso BDTs nd to b dsignd such that bst systm pfomanc is achivd. Ths us may not b th bst fo individua BDTs. III. Dsign of BDT Basd Intactiv Systms In this sction, w focus on BDT basd dcision fusion systms that invov two-way communication and wh a systm componnts a jointy dsignd. Th quaity of such a dcision fusion systm is givn by th cognition at of ach objct and th associatd avag numb of stps in a cognition opation. Th fist quantity svs as a masu of ffctivnss of th fusion systm in caying out dtction/cognition, whi th scond quantity indicats th fficincy in tms of computationa tim/ffot. W popos a mthod of dsigning BDTs, snso us and fusion us. W assum that th snso obsvations a conditionay indpndnt givn th objct cass. W aso assum that ach snso obsvation is chaactizd by a pobabiity distibution function givn ach objct cass and that it is known a pioi. Finay, ach snso uss a th avaiab fatus at ach nod of its BDT. Fist, w popos a way to coopativy us BDTs at th fusion cnt and th snsos. Th fusion cnt BDT is usd to cay out a squntia patition of th objct spac. At ach intna nod of its BDT, th fusion cnt tsts on subst of objcts against anoth subst of objcts. It cocts snso oca dcisions and uss thm to mak a goba dcision on which subst to tst futh, i.., it scts th path of th t to foow. Basd upon this goba dcision, it chooss an appopiat chid nod at which it tsts two nw substs of objcts. Th fusion cnt pats this pocdu ti it achs a tmina nod wh xacty on objct is ft, and thn dcas this objct to b th unknown objct. W notic that at ach nod of th fusion cnt BDT, th oca dcision fom a snso fcts to what dg this snso distinguishs th two substs of objcts that a und tst. To optimay utiiz th capabiity of this snso, it is ncssay that this snso tst th sam two substs of objcts as th fusion cnt bcaus adding/moving objcts to ths two substs woud mak th oca dcision of this snso ss vant to th cognition task at th fusion cnt. Basd on this fact, w t vy snso BDT patition th objct spac in th sam way as th fusion cnt BDT dos. Fo this ason, a th BDTs may b considd idntica with spct to th objct spac. Howv, sinc snso oca dcisions may not aways ag with th goba dcisions, som snsos may not choos th sam path as th fusion cnt if th snsos us thi own oca dcisions to sct chid nods. If so, th oca dcisions fom ths snsos a ss usfu to th fusion cnt. To mak su that a systm mnts foow th sam path of thi BDTs, a mchanism of coodination is ncssay. Such a mchanism is impmntd via a simp two-way communication potoco. Suppos th snsos and th fusion cnt aiv at a nod t, th snsos tansmit thi oca dcisions to th fusion cnt. Basd upon ths oca dcisions, th fusion cnt maks a goba dcision and snds it back to th snsos. Thn both

4 th fusion cnt and th snsos us this goba dcision to choos th sam chid nod. Athough a snso BDTs a th sam in th objct spac, thy a gnay diffnt in th fatu spac and snso dcision u spac. Futhmo, ach snso uss th past goba dcision to dtmin which fatus and what snso dcision u to us at a nod. Now t us invstigat how th stuctu of a BDT affcts th cognition at and th avag numb of stps. Th abiity of such a dcision fusion systm in achiving high cognition ats is mo on th actions at th high v nods of its BDTs. To s this, w not that jction of an objct at a nod pohibits any futh cassification and utimat cognition of this objct. Thfo, an o that occus at a high v nod is mo costy than an o that occus at a dscndnt nod. Fom this viwpoint, at ach nod of th BDT on nds to distinguish objcts that appa most dissimia to th snsos. Fo this pupos, on woud constuct a BDT that tsts and distinguishs th pai of most diffnt objcts at ach nod. Howv, th tota numb of intna nods of such a BDT gows astonomicay with th tota numb of objcts bcaus ony two objcts a distinguishd at ach intna nod. Thus, th avag numb of stps may bcom pohibitivy ag. To aviat this budn, on may choos to tst mo than two objcts at ach nod, but this consqunty dcass th dissimiaity btwn th objcts bing tstd and hnc th cognition ats. In th xtm cas, on may choos to tst a th maining objcts at a sing nod. This wi us th ast avag numb of stps but invitaby incas th pobabiity of o. Thfo, w nd to mak a compomis btwn th dissimiaity of objcts bing tstd and th numb of objcts to tst. In th foowing, w popos a BDT constuction mthod that maks such a compomis. Using this mthod, th mo distinct th objcts a, th ai th stag at which thy a tstd. Thus, th quaity of cassification monotonicay dgads aong a path. This povids a natua way of ncoding th goba dcisions into a squnc of binay bits with dcasing significanc. Such a fomat bcoms usfu whn it is dsid to dtmin ony th goup to which an objct bongs but not its xact idntity. Constuction of BDTs In ou mthod, w stat with th oot nod, and thn pat th foowing pocdu fo a nw nods as ong as thy contain mo than on objct. At ach nod w choos th two substs of objcts to distinguish, cat its chid nods and associat with thm th appopiat sts of objcts fo futh cassification. Suppos w a daing with nod t wh a st Λ(t) of objcts main to b futh cassifid. Ou goa is to sct substs Λ and Λ of Λ(t) accoding to a cition that baancs th goa of objct dissimiaity and th numb of objcts to tst. Lt U dnot th cadinaity of which is th numb of objcts to distinguish. U, Lt D(Λ,Λ ) b th dissimiaity btwn substs Λ and Λ which is dfind as D, Λ = min d O,O ( ) ( ) O,O : O,O Λ wh d(, O ) O is an infomation distanc masu btwn objcts O and O. An ovviw of such distanc masus can b found in [8]. Lt D ( n ) = max (, Λ ) D UΛ = n wh D(n) is th bst possib dissimiaity that can b achivd whn n objcts a tstd. It is not difficut to s that D(n) is a monoton dcasing function of n. Duing th constuction of a BDT, it is dsiab to maximiz both D(n) and n. But thy a conficting objctivs and w nd to baanc ths two objctivs. H w maximiz n subjct to α n n D( n) + D( ) () wh α is a pscibd constant. This cition is basd upon th foowing obsvation. Suppos a objcts occu with th sam pobabiity π. In a cntaizd cognition schm, givn {Λ, Λ and U Λ n, th pobabiity of o P (n) can b = boundd by th foowing [8] P n O,O ( ) ( ) P O Λ O Λ wh P (, O ) O is th pobabiity of o whn ony O and O a tstd. Th ight hand sid can b boundd by a cass of infomation distanc masus btwn O and O [8] and w hav ( ) d ( O,O ) D( Λ, Λ ) P n c O Λ O Λ wh c is a constant. c O Λ O Λ Using th bst possib dissimiaity D(n) w hav P ( n) ( n) ( n) c D c n D O Λ O Λ Using th abov bound as an appoximation fo P (n), w hav

5 n ( ) = n D( n) + ( ) P ( n) P ( ) n D Th ft-hand sid fcts th pfomanc oss du to an incas in th numb of objcts to tst. Dnot th maximum toab pfomanc oss by α, w hav th cition (). Somtims, th a mutip pais of {Λ, Λ that cospond to th sam vau of D(n), i.., thy yid th bst possib dissimiaity. In such situations, on may compa two dsigns using th foowing mthod. Fo ach dsign, find th shotst int-objct distanc btwn Λ and Λ, and thn choos th dsign with th sma such distanc. If th is a ti, count th numb of a th objct pais btwn Λ and Λ that ba this distanc, choos th dsign with th sma count. If th is sti a ti, pat th pvious pocdu fo th nxt shotst distanc. If at th nd, th two dsigns a sti tid, andomy choos any on of thm. This mthod is basd on th fact that shot distancs contibut mo to th o than ong distancs. Finay w psnt ou agoithm fo th constuction of BDTs:. St α. Lt N dnot th st of nw nods. Put oot nod t into N. St t as th cunt nod t.. If th cunt nod t contains xacty on objct, mov t fom N, and go to stp 3; othwis us α to find th maxima n, D(n) and th cosponding {Λ, Λ. Cat ft chid nod t, t Λ( t )= Λ( t)- Λ. Cat ight chid nod t, t Λ( t )= Λ( t)-λ. Put nods t, t into N, mov t fom N. 3. If N is mpty, stop; othwis find a mmb of N, st it as th cunt nod t, and thn go to stp. Dsign of th fusion u and th snso us At ach nod t, th snsos mak oca dcisions gading {Λ, Λ. Basd on ths oca dcisions, th fusion cnt maks a goba dcision. Rca that a nod affcts th systm pfomanc mo than any of its dscndnt nods. Thfo at ach nod, w want to dsign th cosponding fusion u and snso dcision u in such a way that th pobabiity of miscassification at that nod is minimizd gadss of what happns at its dscndnt nods. This is basicay a gdy agoithm. Such us can b considd as th soution to a binay hypothsiststing pobm in which objcts in Λ a tstd against objcts in Λ. Sinc BDTs a usd at a th systm mnts, and th optima fusion u and th optima snso us fo th cunt nod dpnd upon past snso dcisions and past goba dcisions. This adds to th compxity of th pobm. It has bn shown [] that und a mid condition this pobm ducs to a convntiona binay dcision fusion pobm, and th optima fusion u and snso us can b dsignd basd on suts avaiab in []. IV. An Examp In this sction, w psnt an xamp to iustat th suts dvopd in th pvious sction. W us th communication potoco dvopd in th pvious sction to coodinat th snsos and th fusion cnt. W wi discuss constuction of th BDT, dsign of th snso us and th fusion u, and pfomanc vauation. W aso compa this dsign with th cntaizd schm, an ad-hoc M-ay dcision fusion schm and th optimum sing snso schm. Lt us consid a dcision-fusion systm consisting of th indpndnt idntica snsos and a fusion cnt. By idntica snsos w man that th snso obsvations hav th sam chaactistics and a th snsos us th sam BDT. This systm is usd to idntify fou quay iky objcts O, O, O 3 and O 4. Each snso obsvation is assumd to b a scaa. Th objcts a psntd by fou vny spacd points on th a in in th snso obsvation spac as shown in Figu 3. Th distanc btwn adjacnt points is assumd to b a. A snso obsvation is couptd by additiv whit Gaussian nois of zo man and unit vaianc. O 4 O 3 O O -.5a -.5a.5a.5a Figu 3: Objct constation Fo th pupos of compaing ou dsign to an ad-hoc M-ay dcision fusion schm that uss bits fo ach snso, w nd a two v BDT that uss on bit at ach v. Sinc th avag numb of stps is qua to in this pobm, w dsign a BDT and th cosponding snso us and th fusion u that maximizs th avag cognition at. T constuction W us th Kuback divgnc to comput th dissimiaity btwn substs of objcts. Sinc th th snsos a idntica and conditionay indpndnt, th dissimiaity is additiv and w hav d O,O = 3K O, O ( i j ) ( i j ) wh K(, O ) O is th Kuback divgnc btwn i j O i and O j using a sing snso obsvation. Futhmo, sinc th noiss a additiv whit Gaussian, th Kuback divgnc is a quadatic function of th Eucidan distanc btwn objcts x

6 K ( O,O ) i j = O O i Th int-objct distancs fo ou pobm a shown in Tab. Tab : Int-objct distancs d (.,.) O O O 3 O 4 j u = u k =d X k nod {Ο 3, Ο 4 vs. {O, O nod nod 3 u = u k =d O 3a a 7a O 3a 3a a O3 a 3a 3a {O 4 vs. {O 3 O 4 O 3 {O vs. {O O O O4 7a a 3a Sinc th dsid BDT has two vs, w nd to vny divid a objcts into two substs at th oot nod. H n=4, and it is not difficut to find out that D(4)=3a. Th a th candidat dsigns Dsign : Λ ={O, O, Λ ={O 3, O 4. Dsign : Λ ={O, O 4, Λ ={O, O 3. Dsign 3: Λ ={O, O 3, Λ ={O, O 4. Using th sction mthod that w dvopd in th pvious sction in cas of tis, w find that th Dsign is th bst. Hnc at th oot nod, w tst Λ ={O 3, O 4 against Λ ={O, O. Th dsign of th two scond-v nods is tivia, hnc is omittd. Th fusion cnt BDT is shown in Figu 4. U,U,U 3 {Ο 3, Ο 4 vs. {O, O nod nod nod 3 O 4 u = {O 4 vs. {O 3 O 3 u = O Figu 4: Fusion cnt BDT {O vs. {O O u = u = Th BDT usd by th kth snso is shown in Figu 5. In Figu 5, d is th binay dcision mad by th kth snso at stag. W not that th kth snso uss th goba dcision to navigat its BDT. u k =d u k =d u k =d u k =d Figu 5: Th kth Snso BDT Snso dcision u and fusion u Th systm taks two stags to idntify an unknown objct. At stag, it tsts {O, O against {O 3, O 4 and thn chooss on of thm say {O, O. At stag, it tsts O against O. Thus w nd to dsign snso us and th fusion u fo ach stag. At ach stag, w minimiz th tota pobabiity of miscassification. Stag At this stag, w hav a binay hypothsis-tsting pobm of {O,O vs. {O 3,O 4. Th snso dcisions u, u and u 3 a sing binay bits. Basd on ths snso dcisions, th fusion cnt maks a binay goba dcision. It is w known [9] that givn fixd snso us, and givn that th snsos a conditionay indpndnt givn th unknown objct, th optima fusion u is th MAP fusion u which can b xpssd as if p ( u ) p ( u ) i k j k i= k= j= 3 k= u = if p ( u ) < p ( u ) i k j k i= k= j= 3 k= wh p i (u k ) is th conditiona pobabiity of dcision u k of th kth snso whn th unknown objct is O i. It is aso w known [] that th ncssay condition fo a snso u to b optima und th conditiona indpndnc assumption is that it is a ikihood atio tst. Sinc ach snso obsvation X k is a Gaussian andom vaiab, a ikihood atio tst suts in a thshod tst and a binay patitioning of th a in x k. Rca that idntica snso us a usd, and on can asiy s that th dcision bounday is at th oigin x k =. With such snso dcision us, th optima fusion u futh simpifis to a majoity u if u + u + u 3 u = if u + u + u 3

7 Stag At this stag, th systm ith tsts O vs. O, o O 3 vs. O 4. Bcaus of th symmty of th objct constation, th snso dcision us and th fusion u at stag, th sut on tsting O vs. O is ssntiay th sam as that fo tsting O 3 vs. O 4. So without oss of gnaity, w ony consid tsting O vs. O. Now th snso dcisions u, u and u 3 a two-bits vctos with th fist bit psnting th snso dcision at stag and th scond bit psnting th snso dcision at stag. Basd on ths vctos, th fusion cnt maks a goba dcision. Simia to th sut at stag, th optima fusion u is 3 3 if p ( u ) p ( u ) k k k= k= u = 3 3 if p ( u ) < p ( u ) k k k= k= wh p i (u k ) is th conditiona pobabiity that th kth snso dcision is u k whn th unknown objct is O i. Again by th sam agumnt as usd fo stag [], th optima snso us a ikihood atio tsts. Such a tst cosponds to a binay patitioning by mans of a thshod ith in gion x k > if th fist bit of u k is, o in gion x k < if th fist bit of u k is. Ths dcision boundais a functions of th int-objct distanc a. W t a ang fom -5dB to +5dB with stp siz qua to.5db. Fo ach vau of a, w comput th optima snso dcision us, and thn xpss th MAP fusion u as a Booan function of th binay snso dcision vctos. This sut is givn in Tab. H u is th goba dcision that th fusion cnt maks at nod 3. Whn u =, O is dcad, and whn u =, O is dcad. u, u and u 3 a th incoming snso dcision vctos. Sinc th fusion cnt chooss {O, O at th oot nod using a majoity fusion u, th a at ast two s out of th fist bit of u, u and u 3. Diffnt pmutations of ths vctos sha th sam nty in this tab. On can s that th fusion u changs with th intobjct distanc a. Such phnomna a makd by adjacnt and fo u in Tab. Fo sma vaus of a (<-5dB), th fusion cnt pfs O () to O () unss th snsos hav stong suppot fo O. This is bcaus O is sandwichd btwn {O 3, O 4 and O, and th is itt oom fo O. Fo ag vaus of a (>4dB), th fusion cnt chooss O () ov O () unss th snsos stongy suppot O, In this situation, bcaus th signa is sufficinty stong, most of th tim th snso dcision vctos bong to th upp fou ntis in Tab. In such cass, th fusion u is ssntiay a majoity u. Fo intmdiat vaus of a (fom -5dB to 4dB), th fusion u xhibits a chang fom th wak signa fom to th stong signa fom. Tab : Stag fusion u: O (u =) vs. O (u =) a (db) (-5, -5) (-4.5, -3) (-.5, 3) 3.5 (4, 5) u u u 3 u Pfomanc vauation Th systm pfomanc is givn by th pobabiity of miscassification P mc. In Figu 6, th P mc of th BDT basd dcision fusion systm is pottd against th int-objct distanc a as a soid cuv. In th sam figu, w aso pot th pfomanc of th optima cntaizd cassifi and th pfomanc of th optima sing snso cassifi. Th optima cntaizd cassifi uss a th aw snso obsvations to mak a on-stag cassification of th unknown objct. It has th smast possib P mc that svs as a ow bound to th P mc of any oth schm. This P mc is pottd as a dottd cuv. Th optima sing snso cassifi uss on snso obsvation to cassify th unknown objct. Th cosponding P mc is pottd as squas in Figu 6. Th BDT basd dcision fusion systm outpfoms th optima sing snso cassifi. Aso it is sighty infio to th optima cntaizd cassifi. This is futh shown in Figu 7 wh th incas in P mc nomaizd by that of th cntaizd cassifi is pottd. Fo instanc, whn a=-db, fo th optima sing snso cassifi, P mc =.6558; fo th optima cntaizd cassifi, P mc =.588; fo th BDT basd systm, P mc = In Figu 7, th BDT basd dcision fusion systm is compad with th ad-hoc dcision fusion systm in which th th snsos a dsignd as idntica optima sing snso cassifis and th MAP fusion u is usd to fus thi dcisions. Fo ach systm, th incas in P mc with spct to th optima cntaizd cassifi is pottd. It is shown that th BDT basd dcision fusion systm is btt than th ad-hoc dcision fusion systm. Fo instanc, whn a=-db, fo th ad-hoc systm, P mc =.647. Th incas of P mc fo th BDT basd systm is -7dB, fo th ad-hoc systm is -5.5dB.

8 mthodoogy was iustatd by mans of an xamp. Many aspcts of this cass of pobms main unsovd and povid a fuitfu aa fo futu sach. Acknowdgmnt Rsach sponsod by Ai Foc Offic of Scintific Rsach, Ai Foc Systms Command, USAF, und Gant No. F Rfncs Figu 6: Pfomanc of th BDT basd dcision fusion systm Figu 7: Compaison of th BDT basd systm with an ad-hoc systm V. Summay W considd a nw cass of dcision fusion pobms that mpoys binay dcision ts at th snsos and/o at th fusion cnt. Such pobms w catgoizd into fou typs accoding to how th binay dcision ts a dsignd and usd by th fusion systm. Vaious aspcts of th dsign of such systms w discussd. W poposd a systmatic dsign mthodoogy fo on such systm. This [] Pamod K. Vashny, Distibutd Dtction and Data Fusion, Sping-Vag Nw Yok, Inc [] R. Viswanathan and P. K. Vashny, "Distibutd Dtction with Mutip Snsos-Pat I: Fundamntas," Poc. IEEE, vo. 85, No., pp , Jan [3] R. S. Bum, S.A. Kassam and H. V. Poo, "Distibutd Dtction with Mutip Snsos: Pat II Advancd Topics," Poc. IEEE, vo. 85, no., pp Jan [4] S. Rasou and D. Landgb, "A Suvy of Dcision T Cassifi Mthodoogy," IEEE Tans. AES, May 99, pp [5] K. Dmibas, Distibutd Snso Data Fusion with Binay Dcision Ts, IEEE Tans. AES, Vo. 33, No., pp. 3-34, Aug [6] B. Dasaathy, "Opationay Efficint Achitctus fo Fusion of Binay-Dcision Snsos in Mutidcision Envionmnt," Opt. Eng., 36(3), pp , Mach 997. [7] X. Zhu, M. Kam and C. Ros,"M-ay Hypothsis Tsting with Binay Loca Dcisions," Poc. CISS, Pincton, Mach 998. [8] M. Bassvi, Distanc Masus fo Signa Pocssing and Pattn Rcognition, Signa Pocssing, 8 (989) pp [9] Z. Chai and P.K. Vashny, "Optima Data Fusion in Mutip Snso Dtction Systms," IEEE Tans. AES, Vo. AES-, No., pp. 98-, Jan., 986. [] R.R. Tnny and N.R. Sand, "Dtction with Distibutd Snsos," IEEE Tans. AES, Vo. AES-7, pp. 98-, Juy 98. [] Q. Zhang, Ph.D. disstation in pogss.

Bayesian Decision Theory

Bayesian Decision Theory Baysian Dcision Thory Baysian Dcision Thory Know probabiity distribution of th catgoris Amost nvr th cas in ra if! Nvrthss usfu sinc othr cass can b rducd to this on aftr som work Do not vn nd training

More information

Extinction Ratio and Power Penalty

Extinction Ratio and Power Penalty Application Not: HFAN-.. Rv.; 4/8 Extinction Ratio and ow nalty AVALABLE Backgound Extinction atio is an impotant paamt includd in th spcifications of most fib-optic tanscivs. h pupos of this application

More information

EE243 Advanced Electromagnetic Theory Lec # 22 Scattering and Diffraction. Reading: Jackson Chapter 10.1, 10.3, lite on both 10.2 and 10.

EE243 Advanced Electromagnetic Theory Lec # 22 Scattering and Diffraction. Reading: Jackson Chapter 10.1, 10.3, lite on both 10.2 and 10. Appid M Fa 6, Nuuth Lctu # V //6 43 Advancd ctomagntic Thoy Lc # Scatting and Diffaction Scatting Fom Sma Obcts Scatting by Sma Dictic and Mtaic Sphs Coction of Scatts Sphica Wav xpansions Scaa Vcto Rading:

More information

E F. and H v. or A r and F r are dual of each other.

E F. and H v. or A r and F r are dual of each other. A Duality Thom: Consid th following quations as an xampl = A = F μ ε H A E A = jωa j ωμε A + β A = μ J μ A x y, z = J, y, z 4π E F ( A = jω F j ( F j β H F ωμε F + β F = ε M jβ ε F x, y, z = M, y, z 4π

More information

GAUSS PLANETARY EQUATIONS IN A NON-SINGULAR GRAVITATIONAL POTENTIAL

GAUSS PLANETARY EQUATIONS IN A NON-SINGULAR GRAVITATIONAL POTENTIAL GAUSS PLANETARY EQUATIONS IN A NON-SINGULAR GRAVITATIONAL POTENTIAL Ioannis Iaklis Haanas * and Michal Hany# * Dpatmnt of Physics and Astonomy, Yok Univsity 34 A Pti Scinc Building Noth Yok, Ontaio, M3J-P3,

More information

Hydrogen atom. Energy levels and wave functions Orbital momentum, electron spin and nuclear spin Fine and hyperfine interaction Hydrogen orbitals

Hydrogen atom. Energy levels and wave functions Orbital momentum, electron spin and nuclear spin Fine and hyperfine interaction Hydrogen orbitals Hydogn atom Engy lvls and wav functions Obital momntum, lcton spin and nucla spin Fin and hypfin intaction Hydogn obitals Hydogn atom A finmnt of th Rydbg constant: R ~ 109 737.3156841 cm -1 A hydogn mas

More information

Physics 111. Lecture 38 (Walker: ) Phase Change Latent Heat. May 6, The Three Basic Phases of Matter. Solid Liquid Gas

Physics 111. Lecture 38 (Walker: ) Phase Change Latent Heat. May 6, The Three Basic Phases of Matter. Solid Liquid Gas Physics 111 Lctu 38 (Walk: 17.4-5) Phas Chang May 6, 2009 Lctu 38 1/26 Th Th Basic Phass of Matt Solid Liquid Gas Squnc of incasing molcul motion (and ngy) Lctu 38 2/26 If a liquid is put into a sald contain

More information

STATISTICAL MECHANICS OF DIATOMIC GASES

STATISTICAL MECHANICS OF DIATOMIC GASES Pof. D. I. ass Phys54 7 -Ma-8 Diatomic_Gas (Ashly H. Cat chapt 5) SAISICAL MECHAICS OF DIAOMIC GASES - Fo monatomic gas whos molculs hav th dgs of fdom of tanslatoy motion th intnal u 3 ngy and th spcific

More information

Lecture 3.2: Cosets. Matthew Macauley. Department of Mathematical Sciences Clemson University

Lecture 3.2: Cosets. Matthew Macauley. Department of Mathematical Sciences Clemson University Lctu 3.2: Costs Matthw Macauly Dpatmnt o Mathmatical Scincs Clmson Univsity http://www.math.clmson.du/~macaul/ Math 4120, Modn Algba M. Macauly (Clmson) Lctu 3.2: Costs Math 4120, Modn Algba 1 / 11 Ovviw

More information

Overview. 1 Recall: continuous-time Markov chains. 2 Transient distribution. 3 Uniformization. 4 Strong and weak bisimulation

Overview. 1 Recall: continuous-time Markov chains. 2 Transient distribution. 3 Uniformization. 4 Strong and weak bisimulation Rcall: continuous-tim Makov chains Modling and Vification of Pobabilistic Systms Joost-Pit Katon Lhstuhl fü Infomatik 2 Softwa Modling and Vification Goup http://movs.wth-aachn.d/taching/ws-89/movp8/ Dcmb

More information

Mid Year Examination F.4 Mathematics Module 1 (Calculus & Statistics) Suggested Solutions

Mid Year Examination F.4 Mathematics Module 1 (Calculus & Statistics) Suggested Solutions Mid Ya Eamination 3 F. Matmatics Modul (Calculus & Statistics) Suggstd Solutions Ma pp-: 3 maks - Ma pp- fo ac qustion: mak. - Sam typ of pp- would not b countd twic fom wol pap. - In any cas, no pp maks

More information

Seidel s Trapezoidal Partitioning Algorithm

Seidel s Trapezoidal Partitioning Algorithm CS68: Geometic Agoithms Handout #6 Design and Anaysis Oigina Handout #6 Stanfod Univesity Tuesday, 5 Febuay 99 Oigina Lectue #7: 30 Januay 99 Topics: Seide s Tapezoida Patitioning Agoithm Scibe: Michae

More information

GRAVITATION 4) R. max. 2 ..(1) ...(2)

GRAVITATION 4) R. max. 2 ..(1) ...(2) GAVITATION PVIOUS AMCT QUSTIONS NGINING. A body is pojctd vtically upwads fom th sufac of th ath with a vlocity qual to half th scap vlocity. If is th adius of th ath, maximum hight attaind by th body

More information

FI 3103 Quantum Physics

FI 3103 Quantum Physics 7//7 FI 33 Quantum Physics Axan A. Iskana Physics of Magntism an Photonics sach oup Institut Tknoogi Banung Schoing Equation in 3D Th Cnta Potntia Hyognic Atom 7//7 Schöing quation in 3D Fo a 3D pobm,

More information

The angle between L and the z-axis is found from

The angle between L and the z-axis is found from Poblm 6 This is not a ifficult poblm but it is a al pain to tansf it fom pap into Mathca I won't giv it to you on th quiz, but know how to o it fo th xam Poblm 6 S Figu 6 Th magnitu of L is L an th z-componnt

More information

* Meysam Mohammadnia Department of Nuclear Engineering, East Tehran Branch, Islamic Azad University, Tehran, Iran *Author for Correspondence

* Meysam Mohammadnia Department of Nuclear Engineering, East Tehran Branch, Islamic Azad University, Tehran, Iran *Author for Correspondence Indian Jouna o Fundanta and ppid Li Scincs ISSN: 65 Onin n Opn ccss, Onin Intnationa Jouna vaiab at www.cibtch.og/sp.d/js///js.ht Vo. S, pp. 7-/Mysa Rsach tic CQUISITION N NLYSIS OF FLUX N CURRENT COEFFICIENTS

More information

CBSE-XII-2013 EXAMINATION (MATHEMATICS) The value of determinant of skew symmetric matrix of odd order is always equal to zero.

CBSE-XII-2013 EXAMINATION (MATHEMATICS) The value of determinant of skew symmetric matrix of odd order is always equal to zero. CBSE-XII- EXAMINATION (MATHEMATICS) Cod : 6/ Gnal Instuctions : (i) All qustions a compulso. (ii) Th qustion pap consists of 9 qustions dividd into th sctions A, B and C. Sction A compiss of qustions of

More information

8 - GRAVITATION Page 1

8 - GRAVITATION Page 1 8 GAVITATION Pag 1 Intoduction Ptolmy, in scond cntuy, gav gocntic thoy of plantay motion in which th Eath is considd stationay at th cnt of th univs and all th stas and th plants including th Sun volving

More information

Combining Subword and State-level Dissimilarity Measures for Improved Spoken Term Detection in NTCIR-11 SpokenQuery&Doc Task

Combining Subword and State-level Dissimilarity Measures for Improved Spoken Term Detection in NTCIR-11 SpokenQuery&Doc Task Combining Subwod and Stat-lvl Dissimilaity Masus fo Impovd Spokn Tm Dtction in NTCIR-11 SpoknQuy&Doc Task ABSTRACT Mitsuaki Makino Shizuoka Univsity 3-5-1 Johoku, Hamamatsu-shi, Shizuoka 432-8561, Japan

More information

Study on the Classification and Stability of Industry-University- Research Symbiosis Phenomenon: Based on the Logistic Model

Study on the Classification and Stability of Industry-University- Research Symbiosis Phenomenon: Based on the Logistic Model Jounal of Emging Tnds in Economics and Managmnt Scincs (JETEMS 3 (1: 116-1 Scholalink sach Institut Jounals, 1 (ISS: 141-74 Jounal jtms.scholalinksach.og of Emging Tnds Economics and Managmnt Scincs (JETEMS

More information

Exact and Approximate Detection Probability Formulas in Fundamentals of Radar Signal Processing

Exact and Approximate Detection Probability Formulas in Fundamentals of Radar Signal Processing Exact and Approximat tction robabiity Formuas in Fundamntas of Radar Signa rocssing Mark A. Richards Sptmbr 8 Introduction Tab 6. in th txt Fundamntas of Radar Signa rocssing, nd d. [], is rproducd bow.

More information

Chapter 4: Algebra and group presentations

Chapter 4: Algebra and group presentations Chapt 4: Algba and goup psntations Matthw Macauly Dpatmnt of Mathmatical Scincs Clmson Univsity http://www.math.clmson.du/~macaul/ Math 4120, Sping 2014 M. Macauly (Clmson) Chapt 4: Algba and goup psntations

More information

GRAVITATION. (d) If a spring balance having frequency f is taken on moon (having g = g / 6) it will have a frequency of (a) 6f (b) f / 6

GRAVITATION. (d) If a spring balance having frequency f is taken on moon (having g = g / 6) it will have a frequency of (a) 6f (b) f / 6 GVITTION 1. Two satllits and o ound a plant P in cicula obits havin adii 4 and spctivly. If th spd of th satllit is V, th spd of th satllit will b 1 V 6 V 4V V. Th scap vlocity on th sufac of th ath is

More information

CDS 101/110: Lecture 7.1 Loop Analysis of Feedback Systems

CDS 101/110: Lecture 7.1 Loop Analysis of Feedback Systems CDS 11/11: Lctu 7.1 Loop Analysis of Fdback Systms Novmb 7 216 Goals: Intoduc concpt of loop analysis Show how to comput closd loop stability fom opn loop poptis Dscib th Nyquist stability cition fo stability

More information

(, ) which is a positively sloping curve showing (Y,r) for which the money market is in equilibrium. The P = (1.4)

(, ) which is a positively sloping curve showing (Y,r) for which the money market is in equilibrium. The P = (1.4) ots lctu Th IS/LM modl fo an opn conomy is basd on a fixd pic lvl (vy sticky pics) and consists of a goods makt and a mony makt. Th goods makt is Y C+ I + G+ X εq (.) E SEK wh ε = is th al xchang at, E

More information

COMPSCI 230 Discrete Math Trees March 21, / 22

COMPSCI 230 Discrete Math Trees March 21, / 22 COMPSCI 230 Dict Math Mach 21, 2017 COMPSCI 230 Dict Math Mach 21, 2017 1 / 22 Ovviw 1 A Simpl Splling Chck Nomnclatu 2 aval Od Dpth-it aval Od Badth-it aval Od COMPSCI 230 Dict Math Mach 21, 2017 2 /

More information

II.3. DETERMINATION OF THE ELECTRON SPECIFIC CHARGE BY MEANS OF THE MAGNETRON METHOD

II.3. DETERMINATION OF THE ELECTRON SPECIFIC CHARGE BY MEANS OF THE MAGNETRON METHOD II.3. DETEMINTION OF THE ELETON SPEIFI HGE Y MENS OF THE MGNETON METHOD. Wok pupos Th wok pupos is to dtin th atio btwn th absolut alu of th lcton chag and its ass, /, using a dic calld agnton. In this

More information

ON SEMANTIC CONCEPT SIMILARITY METHODS

ON SEMANTIC CONCEPT SIMILARITY METHODS 4 ON SEMANTIC CONCEPT SIMILARITY METHODS Lu Yang*, Vinda Bhavsa* and Haold Boly** *Faculty of Comput Scinc, Univsity of Nw Bunswick Fdicton, NB, E3B 5A3, Canada **Institut fo Infomation Tchnology, National

More information

Shor s Algorithm. Motivation. Why build a classical computer? Why build a quantum computer? Quantum Algorithms. Overview. Shor s factoring algorithm

Shor s Algorithm. Motivation. Why build a classical computer? Why build a quantum computer? Quantum Algorithms. Overview. Shor s factoring algorithm Motivation Sho s Algoith It appas that th univs in which w liv is govnd by quantu chanics Quantu infoation thoy givs us a nw avnu to study & tst quantu chanics Why do w want to build a quantu coput? Pt

More information

Coverage and Rate in Cellular Networks with Multi-User Spatial Multiplexing

Coverage and Rate in Cellular Networks with Multi-User Spatial Multiplexing Covag and Rat in Cllula Ntwoks with Multi-Us Spatial Multiplxing Sjith T. Vtil, Kian Kuchi Dpatmnt of Elctical Engining Indian Institut of Tchnology, Hydabad Hydabad, India 55 {p, kkuchi}@iith.ac.in Anilsh

More information

Estimation of a Random Variable

Estimation of a Random Variable Estimation of a andom Vaiabl Obsv and stimat. ˆ is an stimat of. ζ : outcom Estimation ul ˆ Sampl Spac Eampl: : Pson s Hight, : Wight. : Ailin Company s Stock Pic, : Cud Oil Pic. Cost of Estimation Eo

More information

An Elementary Approach to a Model Problem of Lagerstrom

An Elementary Approach to a Model Problem of Lagerstrom An Elmntay Appoach to a Modl Poblm of Lagstom S. P. Hastings and J. B. McLod Mach 7, 8 Abstact Th quation studid is u + n u + u u = ; with bounday conditions u () = ; u () =. This modl quation has bn studid

More information

Impact of Optical Hop Constraints on the IP/MPLS-over-WSON Networks CAPEX Fengqing Liu 1, Jaume Comellas 2

Impact of Optical Hop Constraints on the IP/MPLS-over-WSON Networks CAPEX Fengqing Liu 1, Jaume Comellas 2 Impact of Optica Hop Constaints on th IP/MPLS-ov-WSON Ntwoks CAPEX Fngqing Liu 1, Jaum Comas 2 1 Schoo of Opto-Ectonic Engining, Nanjing Univsity of Posts and Tcommunications, Nanjing, China 2 Dpt. of

More information

Aakash. For Class XII Studying / Passed Students. Physics, Chemistry & Mathematics

Aakash. For Class XII Studying / Passed Students. Physics, Chemistry & Mathematics Aakash A UNIQUE PPRTUNITY T HELP YU FULFIL YUR DREAMS Fo Class XII Studying / Passd Studnts Physics, Chmisty & Mathmatics Rgistd ffic: Aakash Tow, 8, Pusa Road, Nw Dlhi-0005. Ph.: (0) 4763456 Fax: (0)

More information

Fourier transforms (Chapter 15) Fourier integrals are generalizations of Fourier series. The series representation

Fourier transforms (Chapter 15) Fourier integrals are generalizations of Fourier series. The series representation Pof. D. I. Nass Phys57 (T-3) Sptmb 8, 03 Foui_Tansf_phys57_T3 Foui tansfoms (Chapt 5) Foui intgals a gnalizations of Foui sis. Th sis psntation a0 nπx nπx f ( x) = + [ an cos + bn sin ] n = of a function

More information

Keywords: Auxiliary variable, Bias, Exponential estimator, Mean Squared Error, Precision.

Keywords: Auxiliary variable, Bias, Exponential estimator, Mean Squared Error, Precision. IN: 39-5967 IO 9:8 Ctifid Intnational Jounal of Engining cinc and Innovativ Tchnolog (IJEIT) Volum 4, Issu 3, Ma 5 Imovd Exonntial Ratio Poduct T Estimato fo finit Poulation Man Ran Vija Kuma ingh and

More information

ADDITIVE INTEGRAL FUNCTIONS IN VALUED FIELDS. Ghiocel Groza*, S. M. Ali Khan** 1. Introduction

ADDITIVE INTEGRAL FUNCTIONS IN VALUED FIELDS. Ghiocel Groza*, S. M. Ali Khan** 1. Introduction ADDITIVE INTEGRAL FUNCTIONS IN VALUED FIELDS Ghiocl Goza*, S. M. Ali Khan** Abstact Th additiv intgal functions with th cofficints in a comlt non-achimdan algbaically closd fild of chaactistic 0 a studid.

More information

arxiv: v1 [cond-mat.stat-mech] 27 Aug 2015

arxiv: v1 [cond-mat.stat-mech] 27 Aug 2015 Random matix nsmbls with column/ow constaints. II uchtana adhukhan and Pagya hukla Dpatmnt of Physics, Indian Institut of Tchnology, Khaagpu, India axiv:58.6695v [cond-mat.stat-mch] 7 Aug 5 (Datd: Octob,

More information

Extensive Form Games with Incomplete Information. Microeconomics II Signaling. Signaling Examples. Signaling Games

Extensive Form Games with Incomplete Information. Microeconomics II Signaling. Signaling Examples. Signaling Games Extnsiv Fom Gams ith Incomplt Inomation Micoconomics II Signaling vnt Koçksn Koç Univsity o impotant classs o xtnsiv o gams ith incomplt inomation Signaling Scning oth a to play gams ith to stags On play

More information

Propagation of Current Waves along Quasi-Periodical Thin-Wire Structures: Accounting of Radiation Losses

Propagation of Current Waves along Quasi-Periodical Thin-Wire Structures: Accounting of Radiation Losses Intaction Nots Not 6 3 May 6 Popagation of Cunt Wavs aong Quasi-Piodica Thin-Wi Stuctus: Accounting of Radiation Losss Jügn Nitsch and Sgy Tkachnko Otto-von-Guick-Univsity Magdbug Institut fo Fundamnta

More information

CHAPTER 5 CIRCULAR MOTION

CHAPTER 5 CIRCULAR MOTION CHAPTER 5 CIRCULAR MOTION and GRAVITATION 5.1 CENTRIPETAL FORCE It is known that if a paticl mos with constant spd in a cicula path of adius, it acquis a cntiptal acclation du to th chang in th diction

More information

Bianchi Type V Magnetized Anisotropic Dark Energy Models with Constant Deceleration Parameter

Bianchi Type V Magnetized Anisotropic Dark Energy Models with Constant Deceleration Parameter Th fican Rviw of Physics (0 9:005 69 ianchi Typ V Magntizd nisotopic ak Engy Mods with Constant cation Paamt S Kato,* and Y Shaikh patmnt of Mathmatics, SG mavati Univsity, mavati-60, India patmnt of Mathmatics,

More information

A Comparative Study and Analysis of an Optimized Control Strategy for the Toyota Hybrid System

A Comparative Study and Analysis of an Optimized Control Strategy for the Toyota Hybrid System Pag 563 Wol Elctic Vhicl Jounal Vol. 3 - ISSN 3-6653 - 9 AVERE EVS4 Stavang, Noway, May 13-16, 9 A Compaativ Stuy an Analysis of an Optimiz Contol Statgy fo th Toyota Hybi Systm Tho Hofman 1, Thijs Punot

More information

SME 3033 FINITE ELEMENT METHOD. Bending of Prismatic Beams (Initial notes designed by Dr. Nazri Kamsah)

SME 3033 FINITE ELEMENT METHOD. Bending of Prismatic Beams (Initial notes designed by Dr. Nazri Kamsah) Bnding of Prismatic Bams (Initia nots dsignd by Dr. Nazri Kamsah) St I-bams usd in a roof construction. 5- Gnra Loading Conditions For our anaysis, w wi considr thr typs of oading, as iustratd bow. Not:

More information

A STUDY OF PROPERTIES OF SOFT SET AND ITS APPLICATIONS

A STUDY OF PROPERTIES OF SOFT SET AND ITS APPLICATIONS Intnational sach Jounal of Engining and Tchnology IJET -ISSN: 2395-0056 Volum: 05 Issu: 01 Jan-2018 wwwijtnt p-issn: 2395-0072 STDY O POPETIES O SOT SET ND ITS PPLITIONS Shamshad usain 1 Km Shivani 2 1MPhil

More information

Solid state physics. Lecture 3: chemical bonding. Prof. Dr. U. Pietsch

Solid state physics. Lecture 3: chemical bonding. Prof. Dr. U. Pietsch Solid stat physics Lctu 3: chmical bonding Pof. D. U. Pitsch Elcton chag dnsity distibution fom -ay diffaction data F kp ik dk h k l i Fi H p H; H hkl V a h k l Elctonic chag dnsity of silicon Valnc chag

More information

1. Radiation from an infinitesimal dipole (current element).

1. Radiation from an infinitesimal dipole (current element). LECTURE 3: Radiation fom Infinitsimal (Elmntay) Soucs (Radiation fom an infinitsimal dipol. Duality in Maxwll s quations. Radiation fom an infinitsimal loop. Radiation zons.). Radiation fom an infinitsimal

More information

Physics 202, Lecture 5. Today s Topics. Announcements: Homework #3 on WebAssign by tonight Due (with Homework #2) on 9/24, 10 PM

Physics 202, Lecture 5. Today s Topics. Announcements: Homework #3 on WebAssign by tonight Due (with Homework #2) on 9/24, 10 PM Physics 0, Lctu 5 Today s Topics nnouncmnts: Homwok #3 on Wbssign by tonight Du (with Homwok #) on 9/4, 10 PM Rviw: (Ch. 5Pat I) Elctic Potntial Engy, Elctic Potntial Elctic Potntial (Ch. 5Pat II) Elctic

More information

Loss factor for a clamped edge circular plate subjected to an eccentric loading

Loss factor for a clamped edge circular plate subjected to an eccentric loading ndian ounal of Engining & Matials Scincs Vol., Apil 4, pp. 79-84 Loss facto fo a clapd dg cicula plat subjctd to an ccntic loading M K Gupta a & S P Niga b a Mchanical Engining Dpatnt, National nstitut

More information

SUPPLEMENTARY INFORMATION

SUPPLEMENTARY INFORMATION SUPPLMNTARY INFORMATION. Dtmin th gat inducd bgap cai concntation. Th fild inducd bgap cai concntation in bilay gaphn a indpndntly vaid by contolling th both th top bottom displacmnt lctical filds D t

More information

A Heuristic Approach to Detect Feature Interactions in Requirements

A Heuristic Approach to Detect Feature Interactions in Requirements A Huistic Appoach to Dtct Fatu Intactions in Rquimnts Maitta Hisl Janin Souquiès Fakultät fü Infomatik LORIA Univsité Nancy2 Univsität Magdbug B.P. 239 Bâtimnt LORIA D-39016 Magdbug, Gmany F-54506 Vandœuv-ls-Nancy,

More information

Homework 1 Solutions CSE 101 Summer 2017

Homework 1 Solutions CSE 101 Summer 2017 Homewok 1 Soutions CSE 101 Summe 2017 1 Waming Up 1.1 Pobem and Pobem Instance Find the smaest numbe in an aay of n integes a 1, a 2,..., a n. What is the input? What is the output? Is this a pobem o a

More information

Polarized Transmittance-Reflectance Scatterometry Measurements of 2D Trench Dimensions on Phase-Shift Masks

Polarized Transmittance-Reflectance Scatterometry Measurements of 2D Trench Dimensions on Phase-Shift Masks Poaid Tansmittanc-Rctanc Scattomt Masumnts o D Tnch Dimnsions on Phas-Shit Mass John C. am, Aand Ga, Raa How, Stan Chn n& Tchnoog, nc., Santa Caa, CA 9554 Phsics Dpatmnt, Univsit o Caionia at Davis, CA

More information

What Makes Production System Design Hard?

What Makes Production System Design Hard? What Maks Poduction Systm Dsign Had? 1. Things not always wh you want thm whn you want thm wh tanspot and location logistics whn invntoy schduling and poduction planning 2. Rsoucs a lumpy minimum ffctiv

More information

Geometrical Analysis of the Worm-Spiral Wheel Frontal Gear

Geometrical Analysis of the Worm-Spiral Wheel Frontal Gear Gomtical Analysis of th Wom-Spial Whl Fontal Ga SOFIA TOTOLICI, ICOLAE OACEA, VIRGIL TEODOR, GABRIEL FRUMUSAU Manufactuing Scinc and Engining Dpatmnt, Dunaa d Jos Univsity of Galati, Domnasca st., 8000,

More information

Physics 240: Worksheet 15 Name

Physics 240: Worksheet 15 Name Physics 40: Woksht 15 Nam Each of ths poblms inol physics in an acclatd fam of fnc Althouh you mind wants to ty to foc you to wok ths poblms insid th acclatd fnc fam (i.. th so-calld "won way" by som popl),

More information

Midterm Exam. CS/ECE 181B Intro to Computer Vision. February 13, :30-4:45pm

Midterm Exam. CS/ECE 181B Intro to Computer Vision. February 13, :30-4:45pm Nam: Midtm am CS/C 8B Into to Comput Vision Fbua, 7 :-4:45pm las spa ouslvs to th dg possibl so that studnts a vnl distibutd thoughout th oom. his is a losd-boo tst. h a also a fw pags of quations, t.

More information

Merging to ordered sequences. Efficient (Parallel) Sorting. Merging (cont.)

Merging to ordered sequences. Efficient (Parallel) Sorting. Merging (cont.) Efficient (Paae) Soting One of the most fequent opeations pefomed by computes is oganising (soting) data The access to soted data is moe convenient/faste Thee is a constant need fo good soting agoithms

More information

FREQUENCY DETECTION METHOD BASED ON RECURSIVE DFT ALGORITHM

FREQUENCY DETECTION METHOD BASED ON RECURSIVE DFT ALGORITHM FREQUECY DETECTIO METHOD BAED O RECURIE ALGORITHM Katsuyasu akano*, Yutaka Ota*, Hioyuki Ukai*, Koichi akamua*, and Hidki Fujita** *Dpt. of ystms Managmnt and Engining, agoya Institut of Tchnology, Gokiso-cho,

More information

Basic Polyhedral theory

Basic Polyhedral theory Basic Polyhdral thory Th st P = { A b} is calld a polyhdron. Lmma 1. Eithr th systm A = b, b 0, 0 has a solution or thr is a vctorπ such that π A 0, πb < 0 Thr cass, if solution in top row dos not ist

More information

Differential Kinematics

Differential Kinematics Lctu Diffntia Kinmatic Acknowgmnt : Pof. Ouama Khatib, Robotic Laboato, tanfo Univit, UA Pof. Ha Aaa, AI Laboato, MIT, UA Guiing Qution In obotic appication, not on th poition an ointation, but th vocit

More information

Centralized Multi-Node Repair in Distributed Storage

Centralized Multi-Node Repair in Distributed Storage Cntalizd ulti-nod Rpai in Distibutd Stoag awn Zogui, and Zhiying Wang Cnt fo Pvasiv Communications and Computing (CPCC) Univsity of Califonia, Ivin, USA {mzogui,zhiying}@uci.du Abstact In distibutd stoag

More information

Investigation Effect of Outage Line on the Transmission Line for Karbalaa-132Kv Zone in Iraqi Network

Investigation Effect of Outage Line on the Transmission Line for Karbalaa-132Kv Zone in Iraqi Network Intnational Rsach Jounal of Engining and Tchnology (IRJET) -ISSN: - Volum: Issu: Jun - www.ijt.nt p-issn: - Invstigation Effct of Outag on th Tansmission fo Kabalaa-Kv Zon in Iaqi Ntwok Rashid H. AL-Rubayi

More information

Search sequence databases 3 10/25/2016

Search sequence databases 3 10/25/2016 Sarch squnc databass 3 10/25/2016 Etrm valu distribution Ø Suppos X is a random variabl with probability dnsity function p(, w sampl a larg numbr S of indpndnt valus of X from this distribution for an

More information

Collisionless Hall-MHD Modeling Near a Magnetic Null. D. J. Strozzi J. J. Ramos MIT Plasma Science and Fusion Center

Collisionless Hall-MHD Modeling Near a Magnetic Null. D. J. Strozzi J. J. Ramos MIT Plasma Science and Fusion Center Collisionlss Hall-MHD Modling Na a Magntic Null D. J. Stoi J. J. Ramos MIT Plasma Scinc and Fusion Cnt Collisionlss Magntic Rconnction Magntic connction fs to changs in th stuctu of magntic filds, bought

More information

= ρ. Since this equation is applied to an arbitrary point in space, we can use it to determine the charge density once we know the field.

= ρ. Since this equation is applied to an arbitrary point in space, we can use it to determine the charge density once we know the field. Gauss s Law In diffeentia fom D = ρ. ince this equation is appied to an abita point in space, we can use it to detemine the chage densit once we know the fied. (We can use this equation to ve fo the fied

More information

The Matrix Exponential

The Matrix Exponential Th Matrix Exponntial (with xrciss) by D. Klain Vrsion 207.0.05 Corrctions and commnts ar wlcom. Th Matrix Exponntial For ach n n complx matrix A, dfin th xponntial of A to b th matrix A A k I + A + k!

More information

Randomized Routing as a Regularized Solution to the Route Cost Minimization Problem

Randomized Routing as a Regularized Solution to the Route Cost Minimization Problem andomizd outing as a guaizd Soution to th out Cost Minimization Pobm Vadimi Mabuh Nationa Institut of Standads and Thnoogy 1 Buau Div, Stop 892 Gaithsbug, MD 2899-892, USA ABSTACT Oftn a in-stat outing

More information

Sources. My Friends, the above placed Intro was given at ANTENTOP to Antennas Lectures.

Sources. My Friends, the above placed Intro was given at ANTENTOP to Antennas Lectures. ANTENTOP- 01-008, # 010 Radiation fom Infinitsimal (Elmntay) Soucs Fl Youslf a Studnt! Da finds, I would lik to giv to you an intsting and liabl antnna thoy. Hous saching in th wb gav m lots thotical infomation

More information

The theory of electromagnetic field motion. 6. Electron

The theory of electromagnetic field motion. 6. Electron Th thoy of lctomagntic fild motion. 6. Elcton L.N. Voytshovich Th aticl shows that in a otating fam of fnc th magntic dipol has an lctic chag with th valu dpnding on th dipol magntic momnt and otational

More information

Knowledge Creation with Parallel Teams: Design of Incentives and the Role of Collaboration

Knowledge Creation with Parallel Teams: Design of Incentives and the Role of Collaboration Association fo nfomation Systms AS Elctonic Libay (ASL) AMCS 2009 Pocdings Amicas Confnc on nfomation Systms (AMCS) 2009 Knowldg Cation with Paalll Tams: Dsign of ncntivs and th Rol of Collaboation Shanka

More information

Chapter 8: Homomorphisms

Chapter 8: Homomorphisms Chapt 8: Homomophisms Matthw Macauly Dpatmnt of Mathmatical Scincs Clmson Univsity http://www.math.clmson.du/~macaul/ Math 42, Summ I 24 M. Macauly (Clmson) Chapt 8: Homomophisms Math 42, Summ I 24 / 5

More information

CDS 110b: Lecture 8-1 Robust Stability

CDS 110b: Lecture 8-1 Robust Stability DS 0b: Lct 8- Robst Stabilit Richad M. Ma 3 Fba 006 Goals: Dscib mthods fo psnting nmodld dnamics Div conditions fo obst stabilit Rading: DFT, Sctions 4.-4.3 3 Fb 06 R. M. Ma, altch Gam lan: Robst fomanc

More information

CPE702 Algorithm Analysis and Design Week 11 String Processing

CPE702 Algorithm Analysis and Design Week 11 String Processing CPE702 Agorithm Anaysis and Dsign Wk 11 String Procssing Prut Boonma prut@ng.cmu.ac.th Dpartmnt of Computr Enginring Facuty of Enginring, Chiang Mai Univrsity Basd on Sids by M.T. Goodrich and R. Tamassia

More information

u 3 = u 3 (x 1, x 2, x 3 )

u 3 = u 3 (x 1, x 2, x 3 ) Lctur 23: Curvilinar Coordinats (RHB 8.0 It is oftn convnint to work with variabls othr than th Cartsian coordinats x i ( = x, y, z. For xampl in Lctur 5 w mt sphrical polar and cylindrical polar coordinats.

More information

Bounds on the Probability of Misclassification among Hidden Markov Models

Bounds on the Probability of Misclassification among Hidden Markov Models 20 50th IEEE Confrnc on Dcision and Contro and Europan Contro Confrnc (CDC-ECC) Orando FL USA Dcmr 2-5 20 Bounds on th Proaiity of Miscassification among Hiddn Markov Mods Christoforos Krogou and Christoforos

More information

ENGG 1203 Tutorial. Difference Equations. Find the Pole(s) Finding Equations and Poles

ENGG 1203 Tutorial. Difference Equations. Find the Pole(s) Finding Equations and Poles ENGG 03 Tutoial Systms ad Cotol 9 Apil Laig Obctivs Z tasfom Complx pols Fdbac cotol systms Ac: MIT OCW 60, 6003 Diffc Equatios Cosid th systm pstd by th followig diffc quatio y[ ] x[ ] (5y[ ] 3y[ ]) wh

More information

5.61 Fall 2007 Lecture #2 page 1. The DEMISE of CLASSICAL PHYSICS

5.61 Fall 2007 Lecture #2 page 1. The DEMISE of CLASSICAL PHYSICS 5.61 Fall 2007 Lctu #2 pag 1 Th DEMISE of CLASSICAL PHYSICS (a) Discovy of th Elcton In 1897 J.J. Thomson discovs th lcton and masus ( m ) (and inadvtntly invnts th cathod ay (TV) tub) Faaday (1860 s 1870

More information

ریاضیات عالی پیشرفته

ریاضیات عالی پیشرفته ریاضیات عالی پیشرفته Numic Mthods o Enins مدرس دکتر پدرام پیوندی http://www.pdm-pyvndy.com ~ in Aic Equtions ~ Guss Eimintion Chpt 9 http://numicmthods.n.us.du Sovin Systms o Equtions A in qution in n

More information

Frictional effects, vortex spin-down

Frictional effects, vortex spin-down Chapt 4 Fictional ffcts, votx spin-down To undstand spin-up of a topical cyclon it is instuctiv to consid fist th spin-down poblm, which quis a considation of fictional ffcts. W xamin fist th ssntial dynamics

More information

The Matrix Exponential

The Matrix Exponential Th Matrix Exponntial (with xrciss) by Dan Klain Vrsion 28928 Corrctions and commnts ar wlcom Th Matrix Exponntial For ach n n complx matrix A, dfin th xponntial of A to b th matrix () A A k I + A + k!

More information

XFlow: Internet-Scale Extensible Stream Processing

XFlow: Internet-Scale Extensible Stream Processing XFlow: Intnt-Scal Extnsibl Stam Pocssing Olga Papammanouil, Uğu Çtintml, John Jannotti Dpatmnt of Comput Scinc, Bown Univsity {olga, ugu, jj}@cs.bown.du Jun 30, 2008 Abstact Existing stam pocssing systms

More information

Rectification and Depth Computation

Rectification and Depth Computation Dpatmnt of Comput Engnng Unvst of Cafona at Santa Cuz Rctfcaton an Dpth Computaton CMPE 64: mag Anass an Comput Vson Spng 0 Ha ao 4/6/0 mag cosponncs Dpatmnt of Comput Engnng Unvst of Cafona at Santa Cuz

More information

While flying from hot to cold, or high to low, watch out below!

While flying from hot to cold, or high to low, watch out below! STANDARD ATMOSHERE Wil flying fom ot to cold, o ig to low, watc out blow! indicatd altitud actual altitud STANDARD ATMOSHERE indicatd altitud actual altitud STANDARD ATMOSHERE Wil flying fom ot to cold,

More information

A Self-Tuning Proportional-Integral-Derivative Controller for an Autonomous Underwater Vehicle, Based On Taguchi Method

A Self-Tuning Proportional-Integral-Derivative Controller for an Autonomous Underwater Vehicle, Based On Taguchi Method Jounal of Comput Scinc 6 (8): 862-871, 2010 ISSN 1549-3636 2010 Scinc Publications A Slf-Tuning Popotional-Intgal-Divativ Contoll fo an Autonomous Undwat Vhicl, Basd On Taguchi Mthod M. Santhakuma and

More information

Neural Networks The ADALINE

Neural Networks The ADALINE Lat Lctu Summay Intouction to ua to Bioogica uon Atificia uon McCuoch an itt LU Ronbatt cton Aan Bnaino, a@i.it.ut.t Machin Laning, 9/ ua to h ADALI M A C H I L A R I G 9 / cton Limitation cton aning u

More information

On Self-Avoiding Walks across n-dimensional Dice and Combinatorial Optimization: An Introduction

On Self-Avoiding Walks across n-dimensional Dice and Combinatorial Optimization: An Introduction On Sf-Avoiing Waks acoss n-dimnsiona Dic an ombinatoia Optimization: An Intoction Fanc Bgz ompt Scinc Raigh, N, USA Vsion: Sn Sp 5:5: EDT 3 ontnts Th fab -- Abot Gt an Hans saching fo ks (an abot Jok hiing

More information

and integrated over all, the result is f ( 0) ] //Fourier transform ] //inverse Fourier transform

and integrated over all, the result is f ( 0) ] //Fourier transform ] //inverse Fourier transform NANO 70-Nots Chapt -Diactd bams Dlta uctio W d som mathmatical tools to dvlop a physical thoy o lcto diactio. Idal cystals a iiit this, so th will b som iiitis lii about. Usually, th iiit quatity oly ists

More information

Chapter 5. Control of a Unified Voltage Controller. 5.1 Introduction

Chapter 5. Control of a Unified Voltage Controller. 5.1 Introduction Chapt 5 Contol of a Unifid Voltag Contoll 5.1 Intoduction In Chapt 4, th Unifid Voltag Contoll, composd of two voltag-soucd convts, was mathmatically dscibd by dynamic quations. Th spac vcto tansfomation

More information

Indexed Search Tree (Trie)

Indexed Search Tree (Trie) Indxd Sarch Tr (Tri) Fawzi Emad Chau-Wn Tsng Dpartmnt of Computr Scinc Univrsity of Maryand, Cog Park Indxd Sarch Tr (Tri) Spcia cas of tr Appicab whn Ky C can b dcomposd into a squnc of subkys C 1, C

More information

Chapter 9. Optimization: One Choice Variable. 9.1 Optimum Values and Extreme Values

Chapter 9. Optimization: One Choice Variable. 9.1 Optimum Values and Extreme Values RS - Ch 9 - Optimization: On Vaiabl Chapt 9 Optimization: On Choic Vaiabl Léon Walas 8-9 Vildo Fdico D. Pato 88 9 9. Optimum Valus and Etm Valus Goal vs. non-goal quilibium In th optimization pocss, w

More information

NEWTON S THEORY OF GRAVITY

NEWTON S THEORY OF GRAVITY NEWTON S THEOY OF GAVITY 3 Concptual Qustions 3.. Nwton s thid law tlls us that th focs a qual. Thy a also claly qual whn Nwton s law of gavity is xamind: F / = Gm m has th sam valu whth m = Eath and m

More information

International Journal of Industrial Engineering Computations

International Journal of Industrial Engineering Computations Intnational Jounal of Industial Engining Computations 5 (4 65 74 Contnts lists availabl at GowingScinc Intnational Jounal of Industial Engining Computations hompag: www.gowingscinc.com/ijic A nw modl fo

More information

Reliable Messages and Connection Establishment

Reliable Messages and Connection Establishment 26. Rliabl Mssags Rliabl Mssags and Connction Establishmnt Th attachd pap on liabl mssags is Chapt 10 fom th book Distibutd Systms: Achitctu and Implmntation, ditd by Sap Mullnd, Addison-Wsly, 199. It

More information

Computing and Communications -- Network Coding

Computing and Communications -- Network Coding 89 90 98 00 Computing and Communications -- Ntwork Coding Dr. Zhiyong Chn Institut of Wirlss Communications Tchnology Shanghai Jiao Tong Univrsity China Lctur 5- Nov. 05 0 Classical Information Thory Sourc

More information

Section 6.1. Question: 2. Let H be a subgroup of a group G. Then H operates on G by left multiplication. Describe the orbits for this operation.

Section 6.1. Question: 2. Let H be a subgroup of a group G. Then H operates on G by left multiplication. Describe the orbits for this operation. MAT 444 H Barclo Spring 004 Homwork 6 Solutions Sction 6 Lt H b a subgroup of a group G Thn H oprats on G by lft multiplication Dscrib th orbits for this opration Th orbits of G ar th right costs of H

More information

217Plus TM Integrated Circuit Failure Rate Models

217Plus TM Integrated Circuit Failure Rate Models T h I AC 27Plu s T M i n t g at d c i c u i t a n d i n d u c to Fa i lu at M o d l s David Nicholls, IAC (Quantion Solutions Incoatd) In a pvious issu o th IAC Jounal [nc ], w povidd a highlvl intoduction

More information

A Brief Introduction to Markov Chains and Hidden Markov Models

A Brief Introduction to Markov Chains and Hidden Markov Models A Brief Introduction to Markov Chains and Hidden Markov Modes Aen B MacKenzie Notes for December 1, 3, &8, 2015 Discrete-Time Markov Chains You may reca that when we first introduced random processes,

More information

Monte Carlo Simulation of an Optical Differential Phase-Shift Keying Communication System with Direct Detection Impaired by In-Band Crosstalk

Monte Carlo Simulation of an Optical Differential Phase-Shift Keying Communication System with Direct Detection Impaired by In-Band Crosstalk SIMUL 2012 : Th Fouth Intnational Confnc on Advancs in Systm Simulation Mont Calo Simulation of an Optical Diffntial Phas-Shift Kying Communication Systm with Dict Dtction Impaid y In-Band Cosstalk Gnádio

More information