Identifying Linear Combinations of Ridge Functions

Size: px
Start display at page:

Download "Identifying Linear Combinations of Ridge Functions"

Transcription

1 Advaces Appled Matheatcs 22, Ž Atcle ID aaa , avalable ole at o Idetfyg Lea Cobatos of Rdge Fuctos Mat D. Buha Matheatk, Lehstuhl 8, Uestat Dotud, Dotud, Geay ad Alla Pkus Depatet of Matheatcs, Techo-Isael Isttute of Techology, Hafa, 32000, Isael Receved July 1, 1997; accepted Septebe 19, 1997 Ths pape s about a vese poble. We assue we ae gve a fucto fž. x whch s soe su of dge fuctos of the fo Ý g Ža x. 1 ad we just kow a uppe boud o. We seek to detfy the fuctos g ad also to detfy the dectos a fo such lted foato. Seveal ways to solve ths olea poble ae dscussed ths wok Acadec Pess 1. INTRODUCTION A dge fucto s a ultvaate fucto of the sple fo h:, hž x 1,..., x. gž a1x1 ax. g Ž a x., whee g: ad a a,...,a I othe wods, t s a ultvaate fucto costat o the paallel hypeplaes a x c, c. The vecto a 04 s geeally called the decto. Rdge fuctos appea vaous aeas ad ude vaous guses. We fd the the aea of patal dffeetal equatos Žwhee they have E-al: db@ath.u-dotud.de. E-al: pkus@ath.techo.ac.l $30.00 Copyght 1999 by Acadec Pess All ghts of epoducto ay fo eseved.

2 104 BUHMANN AND PINKUS bee kow fo ay, ay yeas ude the ae of plae waes. 7. We also fd the used coputezed toogaphy 10, statstcs Žwhee they appea pojecto pusut algoths. 5, eual etwoks, ad of couse appoxato theoy. Moe about dge fuctos ay be foud Pkus 11, ad efeeces thee. Whe dealg wth dge fuctos, oe s geeally teested oe of thee possble sets of fuctos. The fst s gve by 1 ½ Ý 5 1 RŽ a,...,a. g Ž a x.: g CŽ., 1,...,. That s, we fx a fte ube of dectos ad we cosde lea cobatos of dge fuctos wth these dectos. The fuctos g ae the vaables. Ths s a lea space. The secod set s ½ Ý R g a x : a 0, g C, 1,...,. Hee, we fx ad we choose both the fuctos g ad the dectos a. Ths s ot a lea space. The thd set s otvated by a odel eual etwoks. It s a subset of the secod. We fx CŽ., called the tasfe fucto eual etwok lteatue, ad we let ½ Ý N c a x b : a 0, c, b, 1,...,. Hee we also fx ad choose both the dectos a Ž called the weghts., ad the shfts b Ž called the thesholds.. Ths s ot a lea space. Oe poble et wth whe dealg wth fucto sets such as the pecedg, s kowg f ad whe a gve fucto s the set. That s, do we have ay way of kowg whethe ay pescbed f s RŽa 1,...,a. fo soe gve dectos a 1,...,a,o R,o N? Whle the fst set s lea, the latte two ae ot, ad ths poble s theefoe fa fo tval. A secod uch elated poble s the followg. Assue you kow Ž 1. 1 o suppose that f s R a,...,a fo soe gve dectos a,...,a, o R,o N. How ca we detee the ukows, be they the fuctos, the dectos, o the shfts? ŽIf we kow a ethod fo deteg these ukows, we essetally have a ethod of fdg whethe we ae the appopate set: we assue that we ae the appopate set,

3 LINEAR COMBINATIONS OF RIDGE FUNCTIONS 105 we detfy the ukows, ad the we check whethe the esultg fucto s fact ou ogal fucto.. I ths pape we addess these questos ad we gve a athe geec ethod of asweg the latte questo. That s, we show that the poble ca be solved. A ajo dawback s that ths ethod s athe oe theoetcal tha pactcal, although pcple ou poofs ae costuctve. I a pevous pape 3, we cosdeed a sla ecovey poble. Thee we assued that we wee gve a fucto G: ad a fucto f:, the latte we kew to be of the fo f Ž x. Ý cg j Ž x t j., x, j1 fo soe ukow coeffcets c 4 4 j j1 ad shfts t j j1, whee Ž o a uppe boud o. s kow. The poble was to detfy the coeffcets ad shfts. The techques developed that pape ae used hee. ŽI that pape we also cosdeed N. The esult obtaed Ž Theoe 6 of. 3, whle ot techcally eo, was eagless. Thus ths pape also allows us to edess ths wog.. 2. UNIQUENESS AND SMOOTHNESS Whe epesetg a fucto as a su of dge fuctos, o whe seekg to detfy ts vaous copoets, t s of fudaetal potace to ty to udestad the extet to whch ay epesetato s uque. We theefoe ask, f k l f Ž x. Ý gž a x. Ý hž b x., 1 1 what ca be sad about the g 4 k 4 k 4 1 ad a 1 elatve to the h 1 l ad b 41 l? Fo leaty popetes, ths poble educes to the followg foulato: Assug Ý g Ž a x. 0, 1 o at least a uppe boud o t beg kow, what ca we say about the g? The followg esult s vald.

4 106 BUHMANN AND PINKUS PROPOSITION 1. If s fte, the a ae pawse lealy depedet, the g CŽ., 1,...,, ad Ý g Ž a x. 0, 1 fo all x, the each g s a polyoal of degee at ost 2. Reak. I fact the boud o the degee of the polyoal ca be futhe educed. Poof. The poof of the poposto s eleetay f each of the g les C 1 Žsee Lea 1 Dacos ad Shahshaha. 4. It goes as 4 4 j follows. Fx 1,...,. Fo each j 1,...,, j, let c satsfy c j a j 0 ad c j a 0. Ths s possble because the a ae pawse lealy depedet. Fo geeal c c,...,c, let Now 1 Dc Ý c s. x s s1 Dgax c a g c Ž a x.. Thus, because each g s suffcetly sooth, 0 ŁD c j Ý g a x j1 1 j ž / j Ž 1. Ý Ł c a g a x 1 j1 j j Ž 1. Ł Ž c a. g Ž a x.. j1 j Fo ou choce of the c j, t follows that g Ž 1. a x 0,

5 LINEAR COMBINATIONS OF RIDGE FUNCTIONS 107 fo all x. Ths ples that g Ž 1. Ž t. 0, fo all t, ad g s theefoe a polyoal of degee at ost 2. If the g ae oly CŽ., the the esult eas vald. I fact we ay 1 eve suppose g s just locally tegable,.e., g L Ž. loc. To pove ths, we use soe vey basc deas fo dstbuto theoy. Choose u such that a u b 0, 1,...,. The 0 Ý gž a Ž x tu.. Ý gž a x tb., 1 1 fo evey t ad x. Let D Co the ftely sooth Ž fuctos wth copact suppot whch we shall call test fuctos.. Thus, 1 s a x 0 g a x tb t dt g s ds. Ý Ý H b ž 1 1 b b H Fx 1,...,4 ad let the c j 4 whch s the sae as j1, j / be as the pevous text. The ž / / 1 1 s a x 0 ŁD j c Ý H gž s. ds, ž b b b j1 j ž / 0 g Ž s. ds Ž c a.. Ž s a x 1 Ž 1. H Ł b b b j1 j j Because s a abtay fucto D, t follows that H Ž 1. g Ž s. Ž s. ds 0, fo all D. It s well kow that ths ples that g s a polyoal of degee at ost 2. Fo copleteess, hee s a shot poof. Assue fst fo splcty that g g C ad that H Ž 1. g s s ds 0, 2.2 fo all D. Fo each atual ube k, let k D have suppot 1k,1k, 0, ad H Ž s. ds 1. Thus 4 s a sequece of k k k k1

6 108 BUHMANN AND PINKUS appoxate dettes, the eleets of such a sequece beg chaactezed by copact suppot that shks to 04 wth gowg dex k, oegatvty, ad ut tegal. Let H g Ž t. gž s. Ž t s. ds. k The g C, ad k H Ž 1. Ž1. g k Ž t. gž s. k Ž t s. ds 0, fo each t. Theefoe g k s a polyoal of degee at ost 2. I addto, g k coveges ufoly to g as k o evey fte teval. Theefoe g s also a polyoal of degee at ost 2. 1 If g s ot cotuous but oly L ad satsfes Ž 2.2. loc fo ay test fucto D, we ca take Foue tasfos ad we ca get by Placheel s detty, H ˆŽ g x. x Ž x. dx 0, 1ˆ whee D ad thus ˆ ae stll abtay. Theefoe, the sese of dstbutos, x 1 ˆgŽ x. 0 whch eas ˆgŽ x. 0 eveywhee except at zeo. Thus Theoe poves the esult, aely, that ˆg s a fte lea cobato of the delta fucto ad ts devatves ceteed at the og, the degees beg lted by 2. ŽWe ae usg hee, of couse, that the vese Foue tasfo of the kth devatve of the delta fucto ceteed at zeo s a algebac polyoal of degee k.. Theefoe we have poved that g ust be a polyoal of degee less tha 1, alost eveywhee. Retug to ou g, t thus follows that each of the g s alost eveywhee a polyoal. The pots whee ay of the g ght ot be a polyoal of degee less tha 1 exted, though the e poduct sde g Ža x., to a hypeplae othogoal to a. Because the su ove of these expessos s detcally zeo ad because the dectos a ae utually lealy depedet, t s a staghtfowad cosequece that each g ust fact be a polyoal eveywhee. Reak. Ths questo of uqueess has bee cosdeed by Albet, Sotag, Mallot 2, Sussa 12, ad Feffea 6 fo N, especally the case whee x tah x. If the dectos a ae pawse lealy depedet, the we ca apply Poposto 1. The codto of pawse lea depedece s ot, howeve, atual ths settg. Poposto 1 k

7 LINEAR COMBINATIONS OF RIDGE FUNCTIONS 109 does pet us to educe the poble to studyg whe Ý c Ž a x. b 1 s a polyoal of degee at ost 2. Hee a ozeo a s fxed; we have gouped all tes wth lealy depedet dectos. We ae teested codtos o,, ad b whch the ply that the c ae all zeo. Is Poposto 1 vald wthout ay estcto o the g? Is t tue that f g :, 1,...,, ad the su ove the g Ža x. vashes detcally fo pawse lealy depedet a, the each g s a polyoal of degee at ost 2? The aswe, ufotuately, s o. It s well kow, see Aczel 1, p. 35 that thee exst hghly ocotuous fuctos h: satsfyg hž x y. hž x. hž y., fo all x ad y. These h ae costucted usg Hael bases. Thus, fo exaple, the equato, 0 g1ž x1. g2ž x2. g3ž x1 x2. has hghly opolyoal, ocotuous solutos g g g h The pecedg text togethe wth Poposto 1, begs the followg teestg questo. Naely, f f x Ý g a x, k Ž. l ad f C, do thee the exst g C fo whch f Ž x. Ý g Ž a x.? 1 I addto, what s the elatoshp betwee k ad l? Fo stace, we would wsh the pevous asseto to hold wth l k. The case k 0sof specal teest. Howeve, hee we wll cosde oly k 1 ad we wll pove ude ld assuptos that deed l k ths case. k Ž PROPOSITION 2. Assue that f has the fo 2.3 ad that f C. 1 fo soe k 1. If, addto, g L fo each, the g C k Ž.. loc

8 110 BUHMANN AND PINKUS Poof. We use the c as the poof of Poposto 1. Let be a abtay test fucto. The the expesso ž / H ŁD j c Ý g Ž a Ž x tu.. Ž t. dt j1 1 j H ž / Ł j1 Ž. D j c f x tu t dt 2.4 j s detcal to the ght-had sde of Ž I othe wods, the dstbutoal devatves of f Ž f s a dstbuto as well as a fucto!. of total ode 1 alog the dectos gve by the c j ae detcal to soe dstbutoal devatve of ode 1of g Ž tes ceta costats.. It does ot atte hee that Ž 2.4. s a uvaate tegal although a weak foulato of a devatve of the ultvaate fucto f would eque a tegal agast a ultvaate test fucto. Ths s because f has cotuous devatves ayway. Now, the dstbutoal devatve s equal to the cotuous classcal devatve of f accodg to Theoe , p. 195 because f C 1 Ž. ad theefoe the sae ust be tue fo the ght-had sde. Thus g C 1 ad Ł j1 j Ž 1. Ł j j1 j D j c f x g a x c a, j Ž 1. kž 1. whee all a ad c ae kow. As such g C Ž., k 1. The esult ow follows. A fucto f x, y s of the fo, Ž R a,...,a 2 fž x, y. Ý g Ž a x by., Ž x, y., 1 fo gve Ž a, b., but ukow cotuous g, 1,...,, f ad oly f ž / b a fž x, y. 0, Ž 3.1. x y Ł 1 a dstbutoal sese. Ths athe sple esult s based o the sae esult the case 1 whch s staghtfowad. Note that ths povdes a

9 LINEAR COMBINATIONS OF RIDGE FUNCTIONS 111 ethod of detfyg RŽa 1,...,a. f 2. Ufotuately a sple chaactezato such as Ž 3.1. does ot hold the case of thee o oe vaables. How ca we detee f a fucto f Ždefed o. s of the fo fo soe gve a,...,a 0 4, but ukow cotuous g,..., g? That s, how do we chaacteze RŽa 1,...,a. 1? Oe aswe ay be foud L ad Pkus 9. Let PŽa 1,...,a. deote the set of polyoals whch vash o all the les a : 4, 1,...,. Ths s a deal. THEOREM 3 ŽL ad Pkus. 9. The cotuous fucto f RŽa 1,...,a. f ad oly f 4 1 f spa q: q polyoal, p D q 0 fo evey p P a,...,a. Ths theoe, ad of tself, does ot povde a sple ethod of checkg whethe a patcula fucto s RŽa 1,...,a.. Howeve, t follows fo the theoy of polyoal deals that oe eed ot check evey p PŽa 1,...,a.. It s a cosequece of Hlbet s bass theoe, c.f. 13, p. 18, that t suffces to cosde ay set of p PŽa 1,...,a. whch ae geeatos fo the polyoal deal PŽa 1,...,a.. Sets of geeatos ae hghly ouque. But t s ot dffcult to show that thee always exsts a faly sple set of geeatos of cadalty. We shall, howeve, ot futhe pusue these deas hee. I Dacos ad Shahshaha 4 ae to be foud two addtoal theoes whch chaacteze RŽa 1,...,a.. The fst s stated oly fo the case 3. But t s fact vald fo evey. 1 THEOREM 4 Dacos ad Shahshaha 4. Let a,...,a be pawse lealy depedet ectos. Let deote the hypeplae c : c a 0, 4 1,...,. A fucto f C Ž. has the fo, f x Ý g a x P x, fo soe polyoal P of degee less tha, f ad oly f Ł c 1 D f 0, fo all choces of c, 1,...,. The fee polyoal te Ž 3.2. s a cosequece of Poposto 1. Aothe uch oe coplcated chaactezato of RŽa 1,...,a. fo ceta patcula choces of Ž due to Royde. ay be foud at the ed of the pape by Dacos ad Shahshaha 4.

10 112 BUHMANN AND PINKUS Ou goal hee s oe odest. Assue we ae gve f RŽa 1,...,a. of the fo Ž We wsh to detfy the g. If f s of the fo 2.3 wth pawse lealy depedet a, the, fo Poposto 1, these g ae uque up to polyoals of degee at ost 2. How do we detee these g? Hee s a ecpe based o the deas we have aleady dscussed. Assue fst that f C 1 Ž. s of the fo Ž Fx 1,...,4 ad let the c j 4j1, j, be as the poof of Poposto 1. I patcula, c j a j 0 ad c j a 0. The we have ŁD jf x D j c Ł c Ý g a x j1 j1 1 j j ž / j Ž 1. Ý Ł c a g a x 1 j1 j ž / j Ž 1. Ł Ž c a. g Ž a x.. j1 j By assupto we kow f ad thus the left-had sde of the pevous equato. We also kow the costat, Ł j1 j j c a, ad the vecto a tself. Thus ths sple ethod gves us g Ž 1.. I othe wods we ca detee g up to a polyoal of degee 2, ageeet wth Poposto 1. 1 Ths aguet also woks f f ad the g ae just L Ž. loc ad 1 L Ž. loc, espectvely. Fo ths, we ague as the poof of Poposto 1;.e., we fd a u as that poof ad we tegate agast a test fucto 1 D. Let ths test fucto scaled by b,.e., Žb 1., be a eleet fo a sequece of appoxate dettes. I patcula, H 1 b. Now, we ca detfy fo the dsplay Ž 2.1. the Ž 1. st devatve of g tegated agast that. Deote the esult by g Ž 1.. So we have 1 Ž 1. t a x j Ž 1. Ł H Ž c a. g Ž t. dt b ž b b, / j1 j 1 Ž 1. Ł j, j1 j g Ž a x. Ž c a., b ž /

11 LINEAR COMBINATIONS OF RIDGE FUNCTIONS 113 Ž 1. whch defes the g. I othe wods, tegatg Ž 1., tes gves g,, aely, the g soothed by the test fucto the afoeetoed way,.e., also scalg by b 1, / t a x g t dt. H ž b b Lettg the suppot of the appoxate detty ted to 04 whle atag ut tegal gves g Ža x. odulo a polyoal of degee 2. The latte s a cosequece of ou Ž 1. -fold tegato of the devatve. Because we ca do ths fo each 1,..., 4, we kow that f s cotaed the set of fuctos, Ý Ž g Ž a x. p Ž a x.., 1 whee each p s a abtay uvaate polyoal of degee at ost 2 ad g s such that g Ž 1. g Ž 1., 1,...,. That s, fo soe choce of polyoals p p we have g g p. Alteatvely, we ay state that f x Ý g a x p x s a ultvaate polyoal of total degee at ost 2 whch ay be deteed ad ca be wtte the fo, pž x. Ý p Ž a x., 1 fo soe choce of p of degee at ost 2. These p Ž ad theefoe the g. ae ot uque. Thee see to be vaous ethods of deteg appopate p. Hee s oe such ethod. We kow that 2 j pž x. Ý p Ž a x. Ý Ý jž a x., 1 1 j0 fo soe choce of coeffcets j. If we ca fd appopate j, we have costucted sutable p, 1,...,. Ths ca be doe the followg Ž. j 4, 2 fasho: Aog the a x 1, j0, choose a bass, ad the wte p tes of ths bass. The coeffcets j fo the polyoal tes of ths bass ca be foud explctly because we kow p. As such we have elucdated oe ecpe fo deteg appopate fuctos g.

12 114 BUHMANN AND PINKUS 4. R Assue f R,.e., f x Ý g a x, fo soe set of cotuous, but ukow, ozeo fuctos g, ad ukow dectos a 0 4, 1,...,. Ca we detee the g ad a? To be oe pecse Ž because thee s a poble of uqueess. we wsh to fd soe g ad a such that Ž 4.1. holds. Fo sooth g, we ca do ths, ude ceta futhe ld assuptos. I ths secto we expla how to fd dectos a 41. We the efe the eade to the pevous secto fo a ecpe fo deteg the g 4 1 based o kowledge of the a 41. The case 1 s elatvely sple. Let us see how t ay be doe because t s stuctve fo the oe coplcated ssues to follow. Assue that f Ž x. g Ž a x., fo soe ukow a Ž a. ad g. Assue also that f Ž 1 ad theefoe g. s cotuously dffeetable. The f x Takg atos we have f x ag a x, 1,...,. Ž x. x f Ž x. x j a,, j 1,...,, a j so log as a ad g Ž a x. j do ot vash. The ght-had sde s depe- det of x fo evey choce of, j 1,..., 4. Note that a ad g ae ot uquely deteed. Specfcally, we ca always eplace a by ca fo ay costat c 0, ad appopately alte g. Thus, kowg all the atos aaj effectvely detees a. Oce gve a, we obta g fo Secto 3. The geealzato to 1 s oe coplcated. We wll use the followg esult, the poof of whch ay be foud Secto 2 of Buha ad Pkus 3, see Theoe 1 ad Coollay 2 thee.

13 LINEAR COMBINATIONS OF RIDGE FUNCTIONS 115 THEOREM 5. Assue that we ae ge ubes b whch satsfy k Ý k 1 k k0 c Ž d. b, k 0,...,21, Ž 4.2. fo ukow ozeo c ad ukow dstct d 1. The d ae uquely deteed as follows. The fucto 0 b0 b-1 b BŽ x. det Ž 4.3. b-1 b2 2 b x x s a polyoal of exact degee. The d, 1,...,, ae ts dstct zeos. The c, 1,...,, ae easly calculated fo the lea equatos Ž 4.2. oce we kow the d, 1,...,. Reak. Note that we ae ot sayg that fo evey choce of b k k0 the syste Ž 4.2. has a soluto. A oe coplete ad detaled exaato of ths poble ay be foud 3. We ow gve a ecpe fo deteg the a 4 Ž , based o vaous Ž ot ueasoable. assuptos. The basc assuptos whch ae used thoughout ae that the a 41 ae pawse dstct, that each g s 2 1 Ž2 1. C soe eghbouhood of 0, ad that g Ž. 0 0, 1,...,. Let c 0 4. Assue that c has bee chose such a way that the values 2 1 Ž2 1. ½ 51 Ž c a. g Ž 0. Ž 4.4. ae ozeo ad dstct. Ths s always possble wth a sutable c, because of the lea depedece of the a. Fo each d 04 ad k 0, 1,..., 2 1, 2 1k k 2 1k k Ž2 1. c d 1 Ý D D f Ž 0. Ž c a. Ž d a. g Ž Ž d a. Ž2 1. Ý Ž c a. g Ž 0.. Ž c a. 1 k

14 116 BUHMANN AND PINKUS If the d a ½ Ž c a. 5 1 Ž 4.5. ae dstct, the t follows fo Theoe 5 that they ay be uquely deteed. Takg lealy depedet d d j, j 1,...,, whch satsfy the pecedg, we obta fo each 1,...,4 the values, d j a ½, j 1,...,. Ž Ž c a. Ths detees the a, 1,...,; Ž c a. Ž Ž. 1. ths case c a. Thus the a ae totally deteed Žsee the eak the case 1.. Ths s ou geeal ecpe. Let us ow cosde ceta of ou equeets futhe detal. ŽSee also the dscusso the poof of Theoe 3 of Buha ad Pkus. 3. We eed to be able to tell whethe, fo a gve d 0 4, the values Ž 4.5. ae dstct. Ths s staghtfowad, because ude ou othe assuptos these values ae dstct f ad oly f the assocated polyoal BŽ x. Ž Ž s of exact degee ad has ths ay dstct zeos. Why do we eed the assupto that the coeffcets Ž 4.4. be dstct, as well as beg ozeo? The dstctess s ot used Theoe 5. It s, howeve, used fo labelg. If two coeffcets have the sae values the we do ot kow how to assg the assocated expessos Ž 4.6. ad thus we detee the a. If the values Ž 4.4. ae ot dstct, we ca alte the c. Fo Theoe 5 we get the values of these coeffcets, ad so t s easly deteed as to whethe these coeffcets ae dstct. ŽThe othe opto s to ty all the dffeet possbltes of assgg the tes Ž 4.6. to the appopate coeffcets, ad the atchg the esults obtaed wth the. Ž2 1. ogal f. Futhe, we ay weake the dead that g Ž. 0 0, 1,...,, by adttg ay shfts away fo the og. Fally, the soothess codto o g ay be weakeed by covolvg g wth a test.e., t detees a up to ultplcato by a ozeo, fte costat fucto ad by cosdeg g, stead of g, as the pevous Ž2 1. secto. Thus the codto g Ž. 0 0 s eplaced by deadg that thee exsts a test fucto such that g og fo all dces. Ž2 1., does ot vash at the

15 LINEAR COMBINATIONS OF RIDGE FUNCTIONS N We ae gve C ad the set N as defed the Itoducto. Because we do ot kow the a, ths s a specfc subset of R. We ca ad wll assue, by usg the ethods of the pevous secto, that we ae able to detfy the a up to ultplcato by costats. That s, a d a fo soe deteed a but udeteed d. Ž Note that we ust hee aga assue that the a ae pawse lealy depedet.. Futheoe we also assue that, usg the ethods of Secto 3, we ca fact detfy g Ž 1. Ž a x. c Ž 1. a x b c Ž 1. d Ž a x. b, fo each. Thus, kowg aleady a, we have educed ou poble to the followg. Gve ad ad c Ž 1. Ž dt b., Ž 5.1. fd the costats c, b, ad d. Ufotuately, we kow of o geeal ethods fo solvg ths poble. Nueous ad hoc ethods peset theselves depedg o the patcula. Fo stace, assue s bouded ad Ž. t whe t, whee. Ž Ths s a typcal case eual etwok applcatos whee usually 0, 1.. Because s bouded, we ca fd c Ž dt b. a, Ž 5.2. fo Ž 5.1. by tegato, whee the ucetaty egadg the polyoal of degee at ost 2 that coes fo the tegato s educed to a costat a Žbecause othe ocostat polyoals ae uled out by boudedess.. Now, lettg t we obta the two values f ad f, oe of whch coespods to c a ad the othe whch coespods to c a, depedg o the sg of d. Thus we obta two optos fo Ž c, a.. If, addto, s stctly ootoe, we ca fd b ad d fo Ž 5.2. by evaluatg at sutable t1 ad t 2. Ths wll gve us oe o two possble choces fo the costats a, b, c, ad d. Howeve, t should be oted that t s vey possble that these costats ae ot uquely deteed. ACKNOWLEDGMENT We thak Joach Stockle ad Bukhad Leze fo potg out the ovesght 3.

16 118 BUHMANN AND PINKUS REFERENCES 1. J. Aczel, Fuctoal Equatos ad The Applcatos, Acadec Pess, New Yok, F. Albet, E. D. Sotag, ad V. Mallot, Uqueess of weghts fo eual etwoks, Atfcal Neual Netwoks fo Speech ad Vso, Ž R. J. Maoe, Ed.., pp , Chapa ad Hall, LodoNew Yok, M. D. Buha ad A. Pkus, O a ecovey poble, A. Nu. Math. 4 Ž 1997., P. Dacos ad M. Shahshaha, O olea fuctos of lea cobatos, SIAM J. Sc. Statst. Coput. 5 Ž 1984., D. L. Dooho ad I. M. Johstoe, Pojecto-based appoxato ad a dualty ethod wth keel ethods, A. Statst. 17 Ž 1989., C. Feffea, Recostuctg a eual et fo ts output, Re. Mat. Ibeoaecaa 10 Ž 1994., F. Joh, Plae Waves ad Sphecal Meas Appled to Patal Dffeetal Equatos, Itescece, New Yok, D. S. Joes, The Theoy of Geealsed Fuctos, Cabdge Uv. Pess, Cabdge, U.K., V. Ya. L ad A. Pkus, Fudaetalty of dge fuctos, J. Appox. Theoy 75 Ž 1993., B. F. Loga ad L. A. Shepp, Optal ecostucto of a fucto fo ts pojectos, Duke Math. J. 42 Ž 1975., A. Pkus, Appoxato by dge fuctos, Suface Fttg ad Multesoluto Methods Ž A. Le Mehaute, C. Rabut, ad L. L. Schuake, Eds.., pp , Vadeblt Uv. Pess, Nashvlle, H. J. Sussa, Uqueess of the weghts fo al feedfowad ets wth a gve putoutput ap, Neual Netwoks 5 Ž 1992., B. L. va de Waede, Modee Algeba Bad II, Spge-Velag, Bel, 1931.

( m is the length of columns of A ) spanned by the columns of A : . Select those columns of B that contain a pivot; say those are Bi

( m is the length of columns of A ) spanned by the columns of A : . Select those columns of B that contain a pivot; say those are Bi Assgmet /MATH 47/Wte Due: Thusday Jauay The poblems to solve ae umbeed [] to [] below Fst some explaatoy otes Fdg a bass of the colum-space of a max ad povg that the colum ak (dmeso of the colum space)

More information

are positive, and the pair A, B is controllable. The uncertainty in is introduced to model control failures.

are positive, and the pair A, B is controllable. The uncertainty in is introduced to model control failures. Lectue 4 8. MRAC Desg fo Affe--Cotol MIMO Systes I ths secto, we cosde MRAC desg fo a class of ult-ut-ult-outut (MIMO) olea systes, whose lat dyacs ae lealy aaetezed, the ucetates satsfy the so-called

More information

FIBONACCI-LIKE SEQUENCE ASSOCIATED WITH K-PELL, K-PELL-LUCAS AND MODIFIED K-PELL SEQUENCES

FIBONACCI-LIKE SEQUENCE ASSOCIATED WITH K-PELL, K-PELL-LUCAS AND MODIFIED K-PELL SEQUENCES Joual of Appled Matheatcs ad Coputatoal Mechacs 7, 6(), 59-7 www.ac.pcz.pl p-issn 99-9965 DOI:.75/jac.7..3 e-issn 353-588 FIBONACCI-LIKE SEQUENCE ASSOCIATED WITH K-PELL, K-PELL-LUCAS AND MODIFIED K-PELL

More information

= y and Normed Linear Spaces

= y and Normed Linear Spaces 304-50 LINER SYSTEMS Lectue 8: Solutos to = ad Nomed Lea Spaces 73 Fdg N To fd N, we eed to chaacteze all solutos to = 0 Recall that ow opeatos peseve N, so that = 0 = 0 We ca solve = 0 ecusvel backwads

More information

Professor Wei Zhu. 1. Sampling from the Normal Population

Professor Wei Zhu. 1. Sampling from the Normal Population AMS570 Pofesso We Zhu. Samplg fom the Nomal Populato *Example: We wsh to estmate the dstbuto of heghts of adult US male. It s beleved that the heght of adult US male follows a omal dstbuto N(, ) Def. Smple

More information

XII. Addition of many identical spins

XII. Addition of many identical spins XII. Addto of may detcal sps XII.. ymmetc goup ymmetc goup s the goup of all possble pemutatos of obects. I total! elemets cludg detty opeato. Each pemutato s a poduct of a ceta fte umbe of pawse taspostos.

More information

χ be any function of X and Y then

χ be any function of X and Y then We have show that whe we ae gve Y g(), the [ ] [ g() ] g() f () Y o all g ()() f d fo dscete case Ths ca be eteded to clude fuctos of ay ube of ado vaables. Fo eaple, suppose ad Y ae.v. wth jot desty fucto,

More information

The Linear Probability Density Function of Continuous Random Variables in the Real Number Field and Its Existence Proof

The Linear Probability Density Function of Continuous Random Variables in the Real Number Field and Its Existence Proof MATEC Web of Cofeeces ICIEA 06 600 (06) DOI: 0.05/mateccof/0668600 The ea Pobablty Desty Fucto of Cotuous Radom Vaables the Real Numbe Feld ad Its Estece Poof Yya Che ad Ye Collee of Softwae, Taj Uvesty,

More information

SUBSEQUENCE CHARACTERIZAT ION OF UNIFORM STATISTICAL CONVERGENCE OF DOUBLE SEQUENCE

SUBSEQUENCE CHARACTERIZAT ION OF UNIFORM STATISTICAL CONVERGENCE OF DOUBLE SEQUENCE Reseach ad Coucatos atheatcs ad atheatcal ceces Vol 9 Issue 7 Pages 37-5 IN 39-6939 Publshed Ole o Novebe 9 7 7 Jyot cadec Pess htt//yotacadecessog UBEQUENCE CHRCTERIZT ION OF UNIFOR TTITIC CONVERGENCE

More information

14. MRAC for MIMO Systems with Unstructured Uncertainties We consider affine-in-control MIMO systems in the form, x Ax B u f x t

14. MRAC for MIMO Systems with Unstructured Uncertainties We consider affine-in-control MIMO systems in the form, x Ax B u f x t Lectue 8 14. MAC o MIMO Systes wth Ustuctued Ucetates We cosde ae--cotol MIMO systes the o, ABu t (14.1) whee s the syste state vecto, u s the cotol put, B s kow costat at, A ad (a dagoal at wth postve

More information

such that for 1 From the definition of the k-fibonacci numbers, the firsts of them are presented in Table 1. Table 1: First k-fibonacci numbers F 1

such that for 1 From the definition of the k-fibonacci numbers, the firsts of them are presented in Table 1. Table 1: First k-fibonacci numbers F 1 Scholas Joual of Egeeg ad Techology (SJET) Sch. J. Eg. Tech. 0; (C):669-67 Scholas Academc ad Scetfc Publshe (A Iteatoal Publshe fo Academc ad Scetfc Resouces) www.saspublshe.com ISSN -X (Ole) ISSN 7-9

More information

On EPr Bimatrices II. ON EP BIMATRICES A1 A Hence x. is said to be EP if it satisfies the condition ABx

On EPr Bimatrices II. ON EP BIMATRICES A1 A Hence x. is said to be EP if it satisfies the condition ABx Iteatoal Joual of Mathematcs ad Statstcs Iveto (IJMSI) E-ISSN: 3 4767 P-ISSN: 3-4759 www.jms.og Volume Issue 5 May. 4 PP-44-5 O EP matces.ramesh, N.baas ssocate Pofesso of Mathematcs, ovt. ts College(utoomous),Kumbakoam.

More information

Recent Advances in Computers, Communications, Applied Social Science and Mathematics

Recent Advances in Computers, Communications, Applied Social Science and Mathematics Recet Advaces Computes, Commucatos, Appled ocal cece ad athematcs Coutg Roots of Radom Polyomal Equatos mall Itevals EFRAI HERIG epatmet of Compute cece ad athematcs Ael Uvesty Cete of amaa cece Pa,Ael,4487

More information

Minimizing spherical aberrations Exploiting the existence of conjugate points in spherical lenses

Minimizing spherical aberrations Exploiting the existence of conjugate points in spherical lenses Mmzg sphecal abeatos Explotg the exstece of cojugate pots sphecal leses Let s ecall that whe usg asphecal leses, abeato fee magg occus oly fo a couple of, so called, cojugate pots ( ad the fgue below)

More information

Some results and conjectures about recurrence relations for certain sequences of binomial sums.

Some results and conjectures about recurrence relations for certain sequences of binomial sums. Soe results ad coectures about recurrece relatos for certa sequeces of boal sus Joha Cgler Faultät für Matheat Uverstät We A-9 We Nordbergstraße 5 Joha Cgler@uveacat Abstract I a prevous paper [] I have

More information

ON THE CONVERGENCE THEOREMS OF THE McSHANE INTEGRAL FOR RIESZ-SPACES-VALUED FUNCTIONS DEFINED ON REAL LINE

ON THE CONVERGENCE THEOREMS OF THE McSHANE INTEGRAL FOR RIESZ-SPACES-VALUED FUNCTIONS DEFINED ON REAL LINE O The Covegece Theoems... (Muslm Aso) ON THE CONVERGENCE THEOREMS OF THE McSHANE INTEGRAL FOR RIESZ-SPACES-VALUED FUNCTIONS DEFINED ON REAL LINE Muslm Aso, Yosephus D. Sumato, Nov Rustaa Dew 3 ) Mathematcs

More information

Chapter 2: Descriptive Statistics

Chapter 2: Descriptive Statistics Chapte : Decptve Stattc Peequte: Chapte. Revew of Uvaate Stattc The cetal teecy of a oe o le yetc tbuto of a et of teval, o hghe, cale coe, ofte uaze by the athetc ea, whch efe a We ca ue the ea to ceate

More information

A Convergence Analysis of Discontinuous Collocation Method for IAEs of Index 1 Using the Concept Strongly Equivalent

A Convergence Analysis of Discontinuous Collocation Method for IAEs of Index 1 Using the Concept Strongly Equivalent Appled ad Coputatoal Matheatcs 27; 7(-): 2-7 http://www.scecepublshggoup.co//ac do:.648/.ac.s.287.2 ISSN: 2328-565 (Pt); ISSN: 2328-563 (Ole) A Covegece Aalyss of Dscotuous Collocato Method fo IAEs of

More information

Fairing of Parametric Quintic Splines

Fairing of Parametric Quintic Splines ISSN 46-69 Eglad UK Joual of Ifomato ad omputg Scece Vol No 6 pp -8 Fag of Paametc Qutc Sples Yuau Wag Shagha Isttute of Spots Shagha 48 ha School of Mathematcal Scece Fuda Uvesty Shagha 4 ha { P t )}

More information

21(2007) Adílson J. V. Brandão 1, João L. Martins 2

21(2007) Adílson J. V. Brandão 1, João L. Martins 2 (007) 30-34 Recuece Foulas fo Fboacc Sus Adílso J. V. Badão, João L. Mats Ceto de Mateátca, Coputa cão e Cog cão, Uvesdade Fedeal do ABC, Bazl.adlso.badao@ufabc.edu.b Depataeto de Mateátca, Uvesdade Fedeal

More information

University of Pavia, Pavia, Italy. North Andover MA 01845, USA

University of Pavia, Pavia, Italy. North Andover MA 01845, USA Iteatoal Joual of Optmzato: heoy, Method ad Applcato 27-5565(Pt) 27-6839(Ole) wwwgph/otma 29 Global Ifomato Publhe (HK) Co, Ltd 29, Vol, No 2, 55-59 η -Peudoleaty ad Effcecy Gogo Gog, Noma G Rueda 2 *

More information

Minimum Hyper-Wiener Index of Molecular Graph and Some Results on Szeged Related Index

Minimum Hyper-Wiener Index of Molecular Graph and Some Results on Szeged Related Index Joual of Multdscplay Egeeg Scece ad Techology (JMEST) ISSN: 359-0040 Vol Issue, Febuay - 05 Mmum Hype-Wee Idex of Molecula Gaph ad Some Results o eged Related Idex We Gao School of Ifomato Scece ad Techology,

More information

Spectral Problems of Two-Parameter System of Operators

Spectral Problems of Two-Parameter System of Operators Pue ad Appled Matheatc Joual 5; 4(4-: 33-37 Publhed ole Augut, 5 (http://wwwcecepublhggoupco//pa do: 648/pa5447 ISSN: 36-979 (Pt; ISSN: 36-98 (Ole Spectal Poble of Two-Paaete Syte of Opeato Rahhada Dhabaadeh

More information

On Eigenvalues of Nonlinear Operator Pencils with Many Parameters

On Eigenvalues of Nonlinear Operator Pencils with Many Parameters Ope Scece Joual of Matheatc ad Applcato 5; 3(4): 96- Publhed ole Jue 5 (http://wwwopececeoleco/oual/oa) O Egevalue of Nolea Opeato Pecl wth May Paaete Rakhhada Dhabaadeh Guay Salaova Depatet of Fuctoal

More information

2. Sample Space: The set of all possible outcomes of a random experiment is called the sample space. It is usually denoted by S or Ω.

2. Sample Space: The set of all possible outcomes of a random experiment is called the sample space. It is usually denoted by S or Ω. Ut: Rado expeet saple space evets classcal defto of pobablty ad the theoes of total ad copoud pobablty based o ths defto axoatc appoach to the oto of pobablty potat theoes based o ths appoach codtoal pobablty

More information

Consider two masses m 1 at x = x 1 and m 2 at x 2.

Consider two masses m 1 at x = x 1 and m 2 at x 2. Chapte 09 Syste of Patcles Cete of ass: The cete of ass of a body o a syste of bodes s the pot that oes as f all of the ass ae cocetated thee ad all exteal foces ae appled thee. Note that HRW uses co but

More information

GREEN S FUNCTION FOR HEAT CONDUCTION PROBLEMS IN A MULTI-LAYERED HOLLOW CYLINDER

GREEN S FUNCTION FOR HEAT CONDUCTION PROBLEMS IN A MULTI-LAYERED HOLLOW CYLINDER Joual of ppled Mathematcs ad Computatoal Mechacs 4, 3(3), 5- GREE S FUCTIO FOR HET CODUCTIO PROBLEMS I MULTI-LYERED HOLLOW CYLIDER Stasław Kukla, Uszula Sedlecka Isttute of Mathematcs, Czestochowa Uvesty

More information

RECAPITULATION & CONDITIONAL PROBABILITY. Number of favourable events n E Total number of elementary events n S

RECAPITULATION & CONDITIONAL PROBABILITY. Number of favourable events n E Total number of elementary events n S Fomulae Fo u Pobablty By OP Gupta [Ida Awad We, +91-9650 350 480] Impotat Tems, Deftos & Fomulae 01 Bascs Of Pobablty: Let S ad E be the sample space ad a evet a expemet espectvely Numbe of favouable evets

More information

A GENERAL CLASS OF ESTIMATORS UNDER MULTI PHASE SAMPLING

A GENERAL CLASS OF ESTIMATORS UNDER MULTI PHASE SAMPLING TATITIC IN TRANITION-ew sees Octobe 9 83 TATITIC IN TRANITION-ew sees Octobe 9 Vol. No. pp. 83 9 A GENERAL CLA OF ETIMATOR UNDER MULTI PHAE AMPLING M.. Ahed & Atsu.. Dovlo ABTRACT Ths pape deves the geeal

More information

On a Problem of Littlewood

On a Problem of Littlewood Ž. JOURAL OF MATHEMATICAL AALYSIS AD APPLICATIOS 199, 403 408 1996 ARTICLE O. 0149 O a Poblem of Littlewood Host Alze Mosbache Stasse 10, 51545 Waldbol, Gemay Submitted by J. L. Bee Received May 19, 1995

More information

Lecture 10: Condensed matter systems

Lecture 10: Condensed matter systems Lectue 0: Codesed matte systems Itoducg matte ts codesed state.! Ams: " Idstgushable patcles ad the quatum atue of matte: # Cosequeces # Revew of deal gas etopy # Femos ad Bosos " Quatum statstcs. # Occupato

More information

Non-degenerate Perturbation Theory

Non-degenerate Perturbation Theory No-degeerate Perturbato Theory Proble : H E ca't solve exactly. But wth H H H' H" L H E Uperturbed egevalue proble. Ca solve exactly. E Therefore, kow ad. H ' H" called perturbatos Copyrght Mchael D. Fayer,

More information

Some Different Perspectives on Linear Least Squares

Some Different Perspectives on Linear Least Squares Soe Dfferet Perspectves o Lear Least Squares A stadard proble statstcs s to easure a respose or depedet varable, y, at fed values of oe or ore depedet varables. Soetes there ests a deterstc odel y f (,,

More information

Inequalities for Dual Orlicz Mixed Quermassintegrals.

Inequalities for Dual Orlicz Mixed Quermassintegrals. Advaces Pue Mathematcs 206 6 894-902 http://wwwscpog/joual/apm IN Ole: 260-0384 IN Pt: 260-0368 Iequaltes fo Dual Olcz Mxed Quemasstegals jua u chool of Mathematcs ad Computatoal cece Hua Uvesty of cece

More information

Non-axial symmetric loading on axial symmetric. Final Report of AFEM

Non-axial symmetric loading on axial symmetric. Final Report of AFEM No-axal symmetc loadg o axal symmetc body Fal Repot of AFEM Ths poject does hamoc aalyss of o-axal symmetc loadg o axal symmetc body. Shuagxg Da, Musket Kamtokat 5//009 No-axal symmetc loadg o axal symmetc

More information

Math 7409 Homework 2 Fall from which we can calculate the cycle index of the action of S 5 on pairs of vertices as

Math 7409 Homework 2 Fall from which we can calculate the cycle index of the action of S 5 on pairs of vertices as Math 7409 Hoewok 2 Fall 2010 1. Eueate the equivalece classes of siple gaphs o 5 vetices by usig the patte ivetoy as a guide. The cycle idex of S 5 actig o 5 vetices is 1 x 5 120 1 10 x 3 1 x 2 15 x 1

More information

= lim. (x 1 x 2... x n ) 1 n. = log. x i. = M, n

= lim. (x 1 x 2... x n ) 1 n. = log. x i. = M, n .. Soluto of Problem. M s obvously cotuous o ], [ ad ], [. Observe that M x,..., x ) M x,..., x ) )..) We ext show that M s odecreasg o ], [. Of course.) mles that M s odecreasg o ], [ as well. To show

More information

MATH 247/Winter Notes on the adjoint and on normal operators.

MATH 247/Winter Notes on the adjoint and on normal operators. MATH 47/Wter 00 Notes o the adjot ad o ormal operators I these otes, V s a fte dmesoal er product space over, wth gve er * product uv, T, S, T, are lear operators o V U, W are subspaces of V Whe we say

More information

Trace of Positive Integer Power of Adjacency Matrix

Trace of Positive Integer Power of Adjacency Matrix Global Joual of Pue ad Appled Mathematcs. IN 097-78 Volume, Numbe 07), pp. 079-087 Reseach Ida Publcatos http://www.publcato.com Tace of Postve Itege Powe of Adacecy Matx Jagdsh Kuma Pahade * ad Mao Jha

More information

2/20/2013. Topics. Power Flow Part 1 Text: Power Transmission. Power Transmission. Power Transmission. Power Transmission

2/20/2013. Topics. Power Flow Part 1 Text: Power Transmission. Power Transmission. Power Transmission. Power Transmission /0/0 Topcs Power Flow Part Text: 0-0. Power Trassso Revsted Power Flow Equatos Power Flow Proble Stateet ECEGR 45 Power Systes Power Trassso Power Trassso Recall that for a short trassso le, the power

More information

Exponential Generating Functions - J. T. Butler

Exponential Generating Functions - J. T. Butler Epoetal Geeatg Fuctos - J. T. Butle Epoetal Geeatg Fuctos Geeatg fuctos fo pemutatos. Defto: a +a +a 2 2 + a + s the oday geeatg fucto fo the sequece of teges (a, a, a 2, a, ). Ep. Ge. Fuc.- J. T. Butle

More information

SUBCLASS OF HARMONIC UNIVALENT FUNCTIONS ASSOCIATED WITH SALAGEAN DERIVATIVE. Sayali S. Joshi

SUBCLASS OF HARMONIC UNIVALENT FUNCTIONS ASSOCIATED WITH SALAGEAN DERIVATIVE. Sayali S. Joshi Faculty of Sceces ad Matheatcs, Uversty of Nš, Serba Avalable at: http://wwwpfacyu/float Float 3:3 (009), 303 309 DOI:098/FIL0903303J SUBCLASS OF ARMONIC UNIVALENT FUNCTIONS ASSOCIATED WIT SALAGEAN DERIVATIVE

More information

A New Method for Solving Fuzzy Linear. Programming by Solving Linear Programming

A New Method for Solving Fuzzy Linear. Programming by Solving Linear Programming ppled Matheatcal Sceces Vol 008 o 50 7-80 New Method for Solvg Fuzzy Lear Prograg by Solvg Lear Prograg S H Nasser a Departet of Matheatcs Faculty of Basc Sceces Mazadara Uversty Babolsar Ira b The Research

More information

GENERALIZATION OF AN IDENTITY INVOLVING THE GENERALIZED FIBONACCI NUMBERS AND ITS APPLICATIONS

GENERALIZATION OF AN IDENTITY INVOLVING THE GENERALIZED FIBONACCI NUMBERS AND ITS APPLICATIONS #A39 INTEGERS 9 (009), 497-513 GENERALIZATION OF AN IDENTITY INVOLVING THE GENERALIZED FIBONACCI NUMBERS AND ITS APPLICATIONS Mohaad Faokh D. G. Depatent of Matheatcs, Fedows Unvesty of Mashhad, Mashhad,

More information

Generalized Duality for a Nondifferentiable Control Problem

Generalized Duality for a Nondifferentiable Control Problem Aeca Joal of Appled Matheatcs ad Statstcs, 4, Vol., No. 4, 93- Avalable ole at http://pbs.scepb.co/aas//4/3 Scece ad Edcato Pblsh DO:.69/aas--4-3 Geealzed Dalty fo a Nodffeetable Cotol Poble. Hsa,*, Vkas

More information

VECTOR MECHANICS FOR ENGINEERS: Vector Mechanics for Engineers: Dynamics. In the current chapter, you will study the motion of systems of particles.

VECTOR MECHANICS FOR ENGINEERS: Vector Mechanics for Engineers: Dynamics. In the current chapter, you will study the motion of systems of particles. Seeth Edto CHPTER 4 VECTOR MECHNICS FOR ENINEERS: DYNMICS Fedad P. ee E. Russell Johsto, J. Systems of Patcles Lectue Notes: J. Walt Ole Texas Tech Uesty 003 The Mcaw-Hll Compaes, Ic. ll ghts eseed. Seeth

More information

Chapter 5 Properties of a Random Sample

Chapter 5 Properties of a Random Sample Lecture 6 o BST 63: Statstcal Theory I Ku Zhag, /0/008 Revew for the prevous lecture Cocepts: t-dstrbuto, F-dstrbuto Theorems: Dstrbutos of sample mea ad sample varace, relatoshp betwee sample mea ad sample

More information

On the limiting law of the length of the longest common and increasing subsequences in random words

On the limiting law of the length of the longest common and increasing subsequences in random words O the ltg law of the legth of the logest coo ad ceasg subsequeces ado wods axv:505.0664v [ath.pr] 0 Sep 06 Jea-Chstophe Beto IRMAR, UMR 665, Uvesté de Rees, 63 Aveue du Gééal Leclec CS 7405, 3504, Rees,

More information

An Enhanced Russell Measure of Super-Efficiency for Ranking Efficient Units in Data Envelopment Analysis

An Enhanced Russell Measure of Super-Efficiency for Ranking Efficient Units in Data Envelopment Analysis Aeca Joual of Appled Sceces 8 (): 92-96, 20 ISSN 546-9239 200 Scece Publcatos A Ehaced Russell Measue of Supe-Effcecy fo Rakg Effcet Uts Data Evelopet Aalyss,2 Al Ashaf,,3 Az B Jaafa,,4 La Soo Lee ad,4

More information

Chapter 7 Varying Probability Sampling

Chapter 7 Varying Probability Sampling Chapte 7 Vayg Pobablty Samplg The smple adom samplg scheme povdes a adom sample whee evey ut the populato has equal pobablty of selecto. Ude ceta ccumstaces, moe effcet estmatos ae obtaed by assgg uequal

More information

chinaxiv: v1

chinaxiv: v1 Matheatcal pcple of essto etwos Zh-Zhog Ta * Zhe Ta. Depatet of physcs, Natog Uvesty, Natog, 69, ha. School of Ifoato Scece ad Techology, Natog Uvesty, Natog, 69, ha (9-3-) Abstact The ufed pocessg ad

More information

Chapter Linear Regression

Chapter Linear Regression Chpte 6.3 Le Regesso Afte edg ths chpte, ou should be ble to. defe egesso,. use sevel mmzg of esdul cte to choose the ght cteo, 3. deve the costts of le egesso model bsed o lest sques method cteo,. use

More information

Distribution of Geometrically Weighted Sum of Bernoulli Random Variables

Distribution of Geometrically Weighted Sum of Bernoulli Random Variables Appled Mathematc, 0,, 8-86 do:046/am095 Publhed Ole Novembe 0 (http://wwwscrpog/oual/am) Dtbuto of Geometcally Weghted Sum of Beoull Radom Vaable Abtact Deepeh Bhat, Phazamle Kgo, Ragaath Naayaachaya Ratthall

More information

Econometric Methods. Review of Estimation

Econometric Methods. Review of Estimation Ecoometrc Methods Revew of Estmato Estmatg the populato mea Radom samplg Pot ad terval estmators Lear estmators Ubased estmators Lear Ubased Estmators (LUEs) Effcecy (mmum varace) ad Best Lear Ubased Estmators

More information

d dt d d dt dt Also recall that by Taylor series, / 2 (enables use of sin instead of cos-see p.27 of A&F) dsin

d dt d d dt dt Also recall that by Taylor series, / 2 (enables use of sin instead of cos-see p.27 of A&F) dsin Learzato of the Swg Equato We wll cover sectos.5.-.6 ad begg of Secto 3.3 these otes. 1. Sgle mache-fte bus case Cosder a sgle mache coected to a fte bus, as show Fg. 1 below. E y1 V=1./_ Fg. 1 The admttace

More information

Optimality Criteria for a Class of Multi-Objective Nonlinear Integer Programs

Optimality Criteria for a Class of Multi-Objective Nonlinear Integer Programs Coucatos Appled Sceces ISSN -77 Volue, Nube,, 7-77 Optalty Ctea o a Class o Mult-Obectve Nolea Itege Pogas Shal Bhagava Depatet o Matheatcs, Babu Shvath Agawal College, Mathua (UP) Ida Abstact hs pape

More information

9. MRAC Design for Affine-in-Control MIMO Systems

9. MRAC Design for Affine-in-Control MIMO Systems Lectue 5 9. MRAC Desg fo Affe--Cotol MIMO Systes Readg ateal: []: Chapte 8, Secto 8.3 []: Chapte 8, Secto 8.5 I ths secto, we cosde MRAC desg fo a class of ult-put-ult-output (MIMO) olea systes whose plat

More information

Council for Innovative Research

Council for Innovative Research Geometc-athmetc Idex ad Zageb Idces of Ceta Specal Molecula Gaphs efe X, e Gao School of Tousm ad Geogaphc Sceces, Yua Nomal Uesty Kumg 650500, Cha School of Ifomato Scece ad Techology, Yua Nomal Uesty

More information

Assignment 5/MATH 247/Winter Due: Friday, February 19 in class (!) (answers will be posted right after class)

Assignment 5/MATH 247/Winter Due: Friday, February 19 in class (!) (answers will be posted right after class) Assgmet 5/MATH 7/Wter 00 Due: Frday, February 9 class (!) (aswers wll be posted rght after class) As usual, there are peces of text, before the questos [], [], themselves. Recall: For the quadratc form

More information

VIII Dynamics of Systems of Particles

VIII Dynamics of Systems of Particles VIII Dyacs of Systes of Patcles Cete of ass: Cete of ass Lea oetu of a Syste Agula oetu of a syste Ketc & Potetal Eegy of a Syste oto of Two Iteactg Bodes: The Reduced ass Collsos: o Elastc Collsos R whee:

More information

CISC 203: Discrete Mathematics for Computing II Lecture 2, Winter 2019 Page 9

CISC 203: Discrete Mathematics for Computing II Lecture 2, Winter 2019 Page 9 Lectue, Wte 9 Page 9 Combatos I ou dscusso o pemutatos wth dstgushable elemets, we aved at a geeal fomula by dvdg the total umbe of pemutatos by the umbe of ways we could pemute oly the dstgushable elemets.

More information

Overview. Review Superposition Solution. Review Superposition. Review x and y Swap. Review General Superposition

Overview. Review Superposition Solution. Review Superposition. Review x and y Swap. Review General Superposition ylcal aplace Soltos ebay 6 9 aplace Eqato Soltos ylcal Geoety ay aetto Mechacal Egeeg 5B Sea Egeeg Aalyss ebay 6 9 Ovevew evew last class Speposto soltos tocto to aal cooates Atoal soltos of aplace s eqato

More information

Born-Oppenheimer Approximation. Kaito Takahashi

Born-Oppenheimer Approximation. Kaito Takahashi o-oppehee ppoato Kato Takahah toc Ut Fo quatu yte uch a ecto ad olecule t eae to ue ut that ft the=tomc UNT Ue a of ecto (ot kg) Ue chage of ecto (ot coulob) Ue hba fo agula oetu (ot kg - ) Ue 4pe 0 fo

More information

2.1.1 The Art of Estimation Examples of Estimators Properties of Estimators Deriving Estimators Interval Estimators

2.1.1 The Art of Estimation Examples of Estimators Properties of Estimators Deriving Estimators Interval Estimators . ploatoy Statstcs. Itoducto to stmato.. The At of stmato.. amples of stmatos..3 Popetes of stmatos..4 Devg stmatos..5 Iteval stmatos . Itoducto to stmato Samplg - The samplg eecse ca be epeseted by a

More information

Harmonic Curvatures in Lorentzian Space

Harmonic Curvatures in Lorentzian Space BULLETIN of the Bull Malaya Math Sc Soc Secod See 7-79 MALAYSIAN MATEMATICAL SCIENCES SOCIETY amoc Cuvatue Loetza Space NEJAT EKMEKÇI ILMI ACISALIOĞLU AND KĀZIM İLARSLAN Aaa Uvety Faculty of Scece Depatmet

More information

Kinematics. Redundancy. Task Redundancy. Operational Coordinates. Generalized Coordinates. m task. Manipulator. Operational point

Kinematics. Redundancy. Task Redundancy. Operational Coordinates. Generalized Coordinates. m task. Manipulator. Operational point Mapulato smatc Jot Revolute Jot Kematcs Base Lks: movg lk fed lk Ed-Effecto Jots: Revolute ( DOF) smatc ( DOF) Geealzed Coodates Opeatoal Coodates O : Opeatoal pot 5 costats 6 paametes { postos oetatos

More information

Best Linear Unbiased Estimators of the Three Parameter Gamma Distribution using doubly Type-II censoring

Best Linear Unbiased Estimators of the Three Parameter Gamma Distribution using doubly Type-II censoring Best Lea Ubased Estmatos of the hee Paamete Gamma Dstbuto usg doubly ype-ii cesog Amal S. Hassa Salwa Abd El-Aty Abstact Recetly ode statstcs ad the momets have assumed cosdeable teest may applcatos volvg

More information

RANDOM SYSTEMS WITH COMPLETE CONNECTIONS AND THE GAUSS PROBLEM FOR THE REGULAR CONTINUED FRACTIONS

RANDOM SYSTEMS WITH COMPLETE CONNECTIONS AND THE GAUSS PROBLEM FOR THE REGULAR CONTINUED FRACTIONS RNDOM SYSTEMS WTH COMPETE CONNECTONS ND THE GUSS PROBEM FOR THE REGUR CONTNUED FRCTONS BSTRCT Da ascu o Coltescu Naval cademy Mcea cel Bata Costata lascuda@gmalcom coltescu@yahoocom Ths pape peset the

More information

Computational Material Chemistry. Kaito Takahashi Institute of Atomic and Molecular Sciences,

Computational Material Chemistry. Kaito Takahashi Institute of Atomic and Molecular Sciences, Coputatoal ateal Chesty Kato Takahash sttute of Atoc ad olecula Sceces kt@gate.sca.edu.tw A Udestad the basc theoy behd quatu chesty calculato Lea ug quatu chesty poga Udestad what the output s sayg Get

More information

1 Solution to Problem 6.40

1 Solution to Problem 6.40 1 Soluto to Problem 6.40 (a We wll wrte T τ (X 1,...,X where the X s are..d. wth PDF f(x µ, σ 1 ( x µ σ g, σ where the locato parameter µ s ay real umber ad the scale parameter σ s > 0. Lettg Z X µ σ we

More information

18.413: Error Correcting Codes Lab March 2, Lecture 8

18.413: Error Correcting Codes Lab March 2, Lecture 8 18.413: Error Correctg Codes Lab March 2, 2004 Lecturer: Dael A. Spelma Lecture 8 8.1 Vector Spaces A set C {0, 1} s a vector space f for x all C ad y C, x + y C, where we take addto to be compoet wse

More information

Stability Analysis for Linear Time-Delay Systems. Described by Fractional Parameterized. Models Possessing Multiple Internal. Constant Discrete Delays

Stability Analysis for Linear Time-Delay Systems. Described by Fractional Parameterized. Models Possessing Multiple Internal. Constant Discrete Delays Appled Mathematcal Sceces, Vol. 3, 29, o. 23, 5-25 Stablty Aalyss fo Lea me-delay Systems Descbed by Factoal Paametezed Models Possessg Multple Iteal Costat Dscete Delays Mauel De la Se Isttuto de Ivestgacó

More information

. The set of these sums. be a partition of [ ab, ]. Consider the sum f( x) f( x 1)

. The set of these sums. be a partition of [ ab, ]. Consider the sum f( x) f( x 1) Chapter 7 Fuctos o Bouded Varato. Subject: Real Aalyss Level: M.Sc. Source: Syed Gul Shah (Charma, Departmet o Mathematcs, US Sargodha Collected & Composed by: Atq ur Rehma (atq@mathcty.org, http://www.mathcty.org

More information

Objectives. Learning Outcome. 7.1 Centre of Gravity (C.G.) 7. Statics. Determine the C.G of a lamina (Experimental method)

Objectives. Learning Outcome. 7.1 Centre of Gravity (C.G.) 7. Statics. Determine the C.G of a lamina (Experimental method) Ojectves 7 Statcs 7. Cete of Gavty 7. Equlum of patcles 7.3 Equlum of g oes y Lew Sau oh Leag Outcome (a) efe cete of gavty () state the coto whch the cete of mass s the cete of gavty (c) state the coto

More information

7.0 Equality Contraints: Lagrange Multipliers

7.0 Equality Contraints: Lagrange Multipliers Systes Optzato 7.0 Equalty Cotrats: Lagrage Multplers Cosder the zato of a o-lear fucto subject to equalty costrats: g f() R ( ) 0 ( ) (7.) where the g ( ) are possbly also olear fuctos, ad < otherwse

More information

Generalized Delta Functions and Their Use in Quasi-Probability Distributions

Generalized Delta Functions and Their Use in Quasi-Probability Distributions Geealzed Delta Fuctos ad The Use Quas-Pobablty Dstbutos RA Bewste ad JD Faso Uvesty of Maylad at Baltmoe Couty, Baltmoe, MD 5 USA Quas-pobablty dstbutos ae a essetal tool aalyzg the popetes of quatum systems,

More information

X ε ) = 0, or equivalently, lim

X ε ) = 0, or equivalently, lim Revew for the prevous lecture Cocepts: order statstcs Theorems: Dstrbutos of order statstcs Examples: How to get the dstrbuto of order statstcs Chapter 5 Propertes of a Radom Sample Secto 55 Covergece

More information

UNIQUENESS IN SEALED HIGH BID AUCTIONS. Eric Maskin and John Riley. Last Revision. December 14, 1996**

UNIQUENESS IN SEALED HIGH BID AUCTIONS. Eric Maskin and John Riley. Last Revision. December 14, 1996** u9d.doc Uqueess u9a/d.doc UNIQUENESS IN SEALED HIGH BID AUCTIONS by Ec Mas ad Joh Rley Last Revso Decembe 4, 996 Depatmet of Ecoomcs, Havad Uvesty ad UCLA A much eale veso of ths pape focussed o the symmetc

More information

NONDIFFERENTIABLE MATHEMATICAL PROGRAMS. OPTIMALITY AND HIGHER-ORDER DUALITY RESULTS

NONDIFFERENTIABLE MATHEMATICAL PROGRAMS. OPTIMALITY AND HIGHER-ORDER DUALITY RESULTS HE PUBLISHING HOUSE PROCEEDINGS OF HE ROMANIAN ACADEMY, See A, OF HE ROMANIAN ACADEMY Volue 9, Nube 3/8,. NONDIFFERENIABLE MAHEMAICAL PROGRAMS. OPIMALIY AND HIGHER-ORDER DUALIY RESULS Vale PREDA Uvety

More information

A Conventional Approach for the Solution of the Fifth Order Boundary Value Problems Using Sixth Degree Spline Functions

A Conventional Approach for the Solution of the Fifth Order Boundary Value Problems Using Sixth Degree Spline Functions Appled Matheatcs, 1, 4, 8-88 http://d.do.org/1.4/a.1.448 Publshed Ole Aprl 1 (http://www.scrp.org/joural/a) A Covetoal Approach for the Soluto of the Ffth Order Boudary Value Probles Usg Sth Degree Sple

More information

22 Nonparametric Methods.

22 Nonparametric Methods. 22 oparametrc Methods. I parametrc models oe assumes apror that the dstrbutos have a specfc form wth oe or more ukow parameters ad oe tres to fd the best or atleast reasoably effcet procedures that aswer

More information

Hájek-Rényi Type Inequalities and Strong Law of Large Numbers for NOD Sequences

Hájek-Rényi Type Inequalities and Strong Law of Large Numbers for NOD Sequences Appl Math If Sc 7, No 6, 59-53 03 59 Appled Matheatcs & Iforato Sceces A Iteratoal Joural http://dxdoorg/0785/as/070647 Háje-Réy Type Iequaltes ad Strog Law of Large Nuers for NOD Sequeces Ma Sogl Departet

More information

The Mathematical Appendix

The Mathematical Appendix The Mathematcal Appedx Defto A: If ( Λ, Ω, where ( λ λ λ whch the probablty dstrbutos,,..., Defto A. uppose that ( Λ,,..., s a expermet type, the σ-algebra o λ λ λ are defed s deoted by ( (,,...,, σ Ω.

More information

A DATA DRIVEN PARAMETER ESTIMATION FOR THE THREE- PARAMETER WEIBULL POPULATION FROM CENSORED SAMPLES

A DATA DRIVEN PARAMETER ESTIMATION FOR THE THREE- PARAMETER WEIBULL POPULATION FROM CENSORED SAMPLES Mathematcal ad Computatoal Applcatos, Vol. 3, No., pp. 9-36 008. Assocato fo Scetfc Reseach A DATA DRIVEN PARAMETER ESTIMATION FOR THE THREE- PARAMETER WEIBULL POPULATION FROM CENSORED SAMPLES Ahmed M.

More information

Multiple Choice Test. Chapter Adequacy of Models for Regression

Multiple Choice Test. Chapter Adequacy of Models for Regression Multple Choce Test Chapter 06.0 Adequac of Models for Regresso. For a lear regresso model to be cosdered adequate, the percetage of scaled resduals that eed to be the rage [-,] s greater tha or equal to

More information

Chapter 9 Jordan Block Matrices

Chapter 9 Jordan Block Matrices Chapter 9 Jorda Block atrces I ths chapter we wll solve the followg problem. Gve a lear operator T fd a bass R of F such that the matrx R (T) s as smple as possble. f course smple s a matter of taste.

More information

Strong Result for Level Crossings of Random Polynomials. Dipty Rani Dhal, Dr. P. K. Mishra. Department of Mathematics, CET, BPUT, BBSR, ODISHA, INDIA

Strong Result for Level Crossings of Random Polynomials. Dipty Rani Dhal, Dr. P. K. Mishra. Department of Mathematics, CET, BPUT, BBSR, ODISHA, INDIA Iteatioal Joual of Reseach i Egieeig ad aageet Techology (IJRET) olue Issue July 5 Available at http://wwwijetco/ Stog Result fo Level Cossigs of Rado olyoials Dipty Rai Dhal D K isha Depatet of atheatics

More information

φ (x,y,z) in the direction of a is given by

φ (x,y,z) in the direction of a is given by UNIT-II VECTOR CALCULUS Dectoal devatve The devatve o a pot ucto (scala o vecto) a patcula decto s called ts dectoal devatve alo the decto. The dectoal devatve o a scala pot ucto a ve decto s the ate o

More information

Atomic units The atomic units have been chosen such that the fundamental electron properties are all equal to one atomic unit.

Atomic units The atomic units have been chosen such that the fundamental electron properties are all equal to one atomic unit. tomc uts The atomc uts have bee chose such that the fudametal electo popetes ae all equal to oe atomc ut. m e, e, h/, a o, ad the potetal eegy the hydoge atom e /a o. D3.33564 0-30 Cm The use of atomc

More information

SOME GEOMETRIC ASPECTS OF VARIATIONAL PROBLEMS IN FIBRED MANIFOLDS

SOME GEOMETRIC ASPECTS OF VARIATIONAL PROBLEMS IN FIBRED MANIFOLDS Electoc tascptos Mathematcal Isttute Slesa Uvesty Opava Czech Republc August Ths tet s a electoc tascpto of the ogal eseach pape D Kupa Some Geometc Aspects of Vaatoal Poblems Fbed Mafolds Fola Fac Sc

More information

Mu Sequences/Series Solutions National Convention 2014

Mu Sequences/Series Solutions National Convention 2014 Mu Sequeces/Seres Solutos Natoal Coveto 04 C 6 E A 6C A 6 B B 7 A D 7 D C 7 A B 8 A B 8 A C 8 E 4 B 9 B 4 E 9 B 4 C 9 E C 0 A A 0 D B 0 C C Usg basc propertes of arthmetc sequeces, we fd a ad bm m We eed

More information

means the first term, a2 means the term, etc. Infinite Sequences: follow the same pattern forever.

means the first term, a2 means the term, etc. Infinite Sequences: follow the same pattern forever. 9.4 Sequeces ad Seres Pre Calculus 9.4 SEQUENCES AND SERIES Learg Targets:. Wrte the terms of a explctly defed sequece.. Wrte the terms of a recursvely defed sequece. 3. Determe whether a sequece s arthmetc,

More information

The Exponentiated Lomax Distribution: Different Estimation Methods

The Exponentiated Lomax Distribution: Different Estimation Methods Ameca Joual of Appled Mathematcs ad Statstcs 4 Vol. No. 6 364-368 Avalable ole at http://pubs.scepub.com/ajams//6/ Scece ad Educato Publshg DOI:.69/ajams--6- The Expoetated Lomax Dstbuto: Dffeet Estmato

More information

A Family of Non-Self Maps Satisfying i -Contractive Condition and Having Unique Common Fixed Point in Metrically Convex Spaces *

A Family of Non-Self Maps Satisfying i -Contractive Condition and Having Unique Common Fixed Point in Metrically Convex Spaces * Advaces Pure Matheatcs 0 80-84 htt://dxdoorg/0436/a04036 Publshed Ole July 0 (htt://wwwscrporg/oural/a) A Faly of No-Self Mas Satsfyg -Cotractve Codto ad Havg Uque Coo Fxed Pot Metrcally Covex Saces *

More information

A New Result On A,p n,δ k -Summabilty

A New Result On A,p n,δ k -Summabilty OSR Joual of Matheatics (OSR-JM) e-ssn: 2278-5728, p-ssn:239-765x. Volue 0, ssue Ve. V. (Feb. 204), PP 56-62 www.iosjouals.og A New Result O A,p,δ -Suabilty Ripeda Kua &Aditya Kua Raghuashi Depatet of

More information

Counting pairs of lattice paths by intersections

Counting pairs of lattice paths by intersections Coutg pas of lattce paths by tesectos Ia Gessel 1, Bades Uvesty, Waltham, MA 02254-9110, USA Waye Goddad 2, Uvesty of Natal, Duba 4000, South Afca Walte Shu, New Yo Lfe Isuace Co., New Yo, NY 10010, USA

More information

Discrete Pseudo Almost Periodic Solutions for Some Difference Equations

Discrete Pseudo Almost Periodic Solutions for Some Difference Equations Advaces Pue Matheatcs 8-7 do:46/ a44 Publshed Ole July (htt://wwwscrpog/joual/a) Dscete Pseudo Alost Peodc Solutos fo Soe Dffeece Equatos Abstact Elhad At Dads * Khall Ezzb Lahce Lhach Uvesty Cad Ayyad

More information

C-1: Aerodynamics of Airfoils 1 C-2: Aerodynamics of Airfoils 2 C-3: Panel Methods C-4: Thin Airfoil Theory

C-1: Aerodynamics of Airfoils 1 C-2: Aerodynamics of Airfoils 2 C-3: Panel Methods C-4: Thin Airfoil Theory ROAD MAP... AE301 Aerodyamcs I UNIT C: 2-D Arfols C-1: Aerodyamcs of Arfols 1 C-2: Aerodyamcs of Arfols 2 C-3: Pael Methods C-4: Th Arfol Theory AE301 Aerodyamcs I Ut C-3: Lst of Subects Problem Solutos?

More information

Part 4b Asymptotic Results for MRR2 using PRESS. Recall that the PRESS statistic is a special type of cross validation procedure (see Allen (1971))

Part 4b Asymptotic Results for MRR2 using PRESS. Recall that the PRESS statistic is a special type of cross validation procedure (see Allen (1971)) art 4b Asymptotc Results for MRR usg RESS Recall that the RESS statstc s a specal type of cross valdato procedure (see Alle (97)) partcular to the regresso problem ad volves fdg Y $,, the estmate at the

More information

Sandwich Theorems for Mcshane Integration

Sandwich Theorems for Mcshane Integration It Joual of Math alyss, Vol 5, 20, o, 23-34 adwch Theoems fo Mcshae Itegato Ismet Temaj Pshta Uvesty Educato Faculty, Pshta, Kosovo temaj63@yahoocom go Tato Taa Polytechc Uvesty Mathematcs Egeeg Faculty,

More information