Local Variability Vector for Text-Independent Speaker Verification

Size: px
Start display at page:

Download "Local Variability Vector for Text-Independent Speaker Verification"

Transcription

1 Loal Vaiabiliy Veo fo ex-inepenen Speake Veifiaion Liping Chen 1 Kong Aik Lee 2 Bin Ma 2 Wu Guo 1 Haizhou Li 2 an Li Rong Dai 1 1 Naional Engineeing Laboaoy fo Speeh an Language Infomaion Poessing Univesiy of Siene an ehnology of China (USC) 2 Insiue fo Infoomm Reseah Ageny fo Siene ehnology an Reseah (A*SAR) lp2011@mail.us.eu.n kalee@i2.a-sa.eu.sg Absa oal vaiabiliy moeling has shon o be effeive fo exinepenen speake veifiaion ask. I povisions a aable ay o esimae he so-alle i-veo hih esibes he speake an session vaiabiliy enee in an ueane. Due o he lo imensionaliy of he i-veo hannel ompensaion ehniques suh as linea isiminan analysis (LDA) an pobabilisi LDA an be applie fo he pupose of hannel ompensaion. his pape poposes he loal vaiabiliy moeling ehnique he enal iea of hih is o apue he loal vaiabiliy assoiae ih iniviual imension of he aousi spae. We analyze he laen suue assoiae ih boh he i-veo an loal vaiabiliy veo an sho ha he o epesenaions omplemen eah ohe base on he expeimen onue on NIS SRE 08 an SRE 10 aases. Inex ems speake eogniion fao analysis session vaiabiliy 1. Inouion Ove he pas fe yeas many appoahes base on he Gaussian mixue moel (GMM) in a GMM-UBM fameok [1] have been popose fo ex-inepenen speake veifiaion ask [2]. Folloing he same fomulaion as in he join fao analysis (JFA) [3] he oal vaiabiliy moel [4] onfines he speake an hannel vaiabiliy ihin a lo-imensional subspae leaing o a fixe an eue imension epesenaion fo speeh ueanes i.e. he so-alle i-veo. eaing an i- veo as a ompa epesenaion of a speeh ueane hannel ompensaion ehniques fo insane ihin-lass ovaiane nomalizaion [5] linea isiminan analysis (LDA) [6] an pobabilisi LDA (PLDA) [7] an hen be applie effeively on he lo-imensional i-veos. An i-veo oul be seen as a eue-imension epesenaion of a GMM mean supeveo (obaine by onaenaing he mean veos in he GMM). hough imension euion oul be pefome on he supeveo using eeminisi ehniques e.g. piniple omponen analysis (PCA) [6] he i-veo exaion is fomulae in pobabilisi ems base on a laen vaiable moel [6]. One obvious benefi is ha in aiion o obaining he i-veo as he poseio mean of he laen vaiable e oul also ompue he poseio ovaiane hih quanifies he uneainy of he esimae an fol in he infomaion in subsequen moeling [8]. Among ohes pobabilisi PCA an fao analysis ae o ommonly use laen vaiable moels in speeh appliaions. he oal vaiabiliy moel an he JFA alike is an exension o he lassial fao analysis ih aiional ying of mixues an fames of an ueane oniione on he same laen vaiable. his ill be fuhe illusae in Seion 2. In his pape e analyze he ying sheme use in he oal vaiabiliy moel (VM) an popose a iffeen appoah by elaxing he ying of mixues suh ha eah iniviual imension of aousi feaue is oniione on is on laen vaiable. A speeh ueane is heeby epesene by a se of loal vaiabiliy veos insea of a single i-veo. We efe o he popose moel as he loal vaiabiliy moel (LVM) he enal iea of hih is o apue he loal vaiabiliy assoiae ih iniviual imension of he aousi spae. In his pape e eive he poseio infeene an give he maximum likelihoo esimae of he moel paamees using he expeaion-maximizaion (EM) algoihm. We also emonsae he use of he popose moel fo ex-inepenen speake veifiaion ask. he es of he pape is oganize as follos. Seion 2 pesens a bief ovevie of he oal vaiabiliy moel an he i-veo exaion poess. Seion 3 poposes he loal vaiabiliy moel an shos ho o use loal vaiabiliy veos fo speake veifiaion ih PLDA. Seion 4 gives some expeimen esuls. Finally Seion 5 onlues he pape. 2. he i-veo paaigm his session gives a bief ovevie of he i-veo exaion poess. In paiula e emphasize on he analysis of laen vaiable ying aoss fames an mixues so as o esablish he onneion o he ne ying sheme popose in his pape I-veo exaion he pupose of i-veo exaion is o fin a fixe-lengh lo-imensional epesenaion fo vaiable-lengh ueanes. he funamenal assumpion of i-veo is ha he feaue veo sequene of an ueane is geneae fom a session-speifi GMM hose mean supeveo m is onsaine o lie in a lo imensional subspae ih oigin μ as follos m μ. (1) Hee is he session inex hee a laen vaiable is assoiae o eah ueane. he oal vaiabiliy maix moels he speake an hannel vaiabiliy leane fom a aining se. An i-veo is hen aken as he poseio mean of he laen vaiable epesening boh he speake an hannel infomaion of an ueane [4].

2 μ Σ 1 C μ V Σ 1 D o o 1 N Figue 1: Pobabilisi gaphial moel illusaing he oal vaiabiliy moel (VM). N is he numbe of fames fom session ha assoiae ih he -h Gaussian omponen hee C is he numbe of Gaussian omponens an R is he numbe of ueanes. An i-veo is aken as he poseio mean of he laen vaiable oal vaiabiliy moel 1 C 1 R Figue 1 shos he oal vaiabiliy moel (VM) in he fom of a pobabilisi gaphial moel. Hee C enoes he numbe of Gaussian omponens hile μ an Σ ae he mean veo an ovaiane maix of he -h Gaussian espeively. In Fig. 1 he obsevaions ae he aousi feaue veos o epesene ih a shae ile. he eangula box suouning he ile ih he value N a is boom igh one iniaes ha hee ae N numbe of obseve veos fom he -h Gaussian fo he -h speeh segmen: po o μ Σ fo N. (2) Noie ha e eompose he oal vaiabiliy maix 1 2 C o is omponen maies one assoiae ih eah Gaussian. In Fig. 1 he oue box iniaes ha he same opeaion is epeae fo all Gaussian omponens fo C. Eah Gaussian omponen aouns fo N numbe of obseve veos o 1 o N N olleively epesene as he union of hih gives ise o he obseve sequene. Fo given a aase he numbe of segmens is enoe as R. One impoan feaue of he oal vaiabiliy moel (an he join fao analysis alike) is ying of he obseve isibuions oniione on he same laen vaiable aoss fames an mixues. Fo he uen ase he ying appeas a o plaes. Fisly he laen vaiable is ie aoss obsevaions o fo 1... N peaining o a Gaussian. Seonly he same laen vaiable is ie aoss he C Gaussian omponens. In Fig. 1 he ying of vaiable is eflee by plaing ousie he o eangula boxes hih essenially iniaes ha he same laen vaiable (un-shae ile) is ie aoss mixues an aoss fames of a given speeh segmen. In mahemaial noaion his is eflee by opping he mixue inex on he laen vaiable. he noion of ying he laen vaiable aoss fames is base on he assumpion ha he hannel an speake being onsan houghou a given speeh segmen (e.g. spoken by Figue 2: Pobabilisi gaphial moel illusaing he loal vaiabiliy moel (LVM). he same imension aoss iffeen Gaussians shae he same loal vaiabiliy veo. he same peson using he same hanse). his assumpion leas o he folloing likelihoo funion: lvm o μ Σ I. 0 (3) Noie ha in (3) he laen vaiable ~ 0 II is assume o follos sana nomal pio. Fuhemoe he same vaiable is ie aoss fames an mixues fo a single session hile sepaae vaiables ae use fo he R speeh segmens o sessions available fo aining. Noie also epesens he se of moel paamees μ Σ; 12 C C hee he mean veos an ovaiane maies ae geneally aken as hose of he UBM. hough i is possible o upae he mean veos an ovaiane maies hey ae usually fixe. Essenially he loaing maies fo 1 2 C C ae he emaining paamees o be opimize. Conaenaing hese maies one afe anohe in a olumn ise manne e fom he so-alle oal vaiabiliy maix. 3. Loal vaiabiliy veo As shon in Seion 2 an i-veo epesens boh he speake an session vaiabiliy onaine in an ueane. o moel he loal vaiabiliy speifi o eah imension of he aousi spae e popose o emove he ying of laen vaiable aoss imensions hile eaining he ying aoss fames an mixues. his fomulaion leas o imension-eni vaiabiliy moeling hih e efe o as he loal vaiabiliy moel. In he folloing e sho ho suh ying sheme oul be aomplishe Loal vaiabiliy moel 1 D 1 R he objeive of loal vaiabiliy moel is o exa he loal vaiabiliy assoiae ih eah iniviual imension of he aousi feaues by eiaing one laen vaiable o evey single imension. Le m m m m D enoes he D 1 veo epesening he mean of he -h Gaussian hee D is he imension of aousi feaue veos. Wih his noaion he mean supeveo m on he lef-han-sie of (1) is obaine by onaenaing he mean veos in he GMM suh ha m m1 m2... m C foms he CD 1 supeveo.

3 o moel he loal vaiabiliy assoiae ih eah iniviual imension of he aousi spae e e-oganize he elemens in he supeveo o fom D sub-veos m m 1 m2... mc fo = 1 2 D eah onsising of C elemens. Essenially he veos m un hough all he C Gaussian omponens of he GMM one imension a a ime. In he above e have oppe he session inex fo beviy. aking he session inex ino aoun e onfine m o lie in a lo-imensional subspae: m m μ V. (4) In (4) he global mean μ is fome in simila manne as m hile V an ae he subspae an laen vaiable apuing he loal vaiabiliy assoiae ih he -h imension of he aousi spae. Figue 2 shos he LVM in he fom of gaphial moel. hee ae o majo iffeenes fom ha of he VM of Fig. 1. Fisly he LVM is imension-eni insea of mixueeni. his is eflee by he hange of subsip fom o a he boom igh one of he seon eangula box. Seonly he ile epesening he laen vaiable is no loae insie he seon eangula box. By his e assign sepaae laen vaiables o he D aousi feaues hile ie aoss mixues an fames of an ueane. Moe speifially he popose ying sheme leas o he folloing likelihoo funion: o o 0. (5) l o V I I LVM he iffeene beeen LVM fom VM an be obseve in hei likelihoo funions in (3) an (5). Noably he laen vaiable in LVM is maginalize sepaaely fo eah imension of he aousi spae as oppose o maginalizaion ove he pou aoss mixues in (3) fo he ase of VM. Like VM he loal vaiabiliy spaes V fo = 1 2 D ae he only moel paamees e nee o esimae. Also he pio of he laen vaiable is assume o follo he sana nomal isibuion. A pe-poessing sep is essenial fo he fomulaion in (5) o be possible. ha is he obseve veo o has o be enalize an hiene in avane hee o epesens he esuling veo. Fom he implemenaion pespeive his oul be aomplishe by enalizing an hiening he fis-oe saisis as esibe in [10] Poseio infeene an paamee esimaion In (5) similaly in (3) e assume ha he alignmen of fames o Gaussian omponens is knon. In paie his infomaion is given by he zeo-oe an fis-oe saisis F exae ih he UBM [3]. In paiula he zeo-oe saisis oul be olleively epesene in a maix fom as Γ (6) C hee iniaes he numbe of fames in ha falls ino he -h Gaussian omponens of he UBM. Given he imension-eni naue of he LVM he poseio mean an ovaiane of he laen vaiable eah assoiae ih an iniviual imension of he aousi spae oul be infee sepaaely as follos: y E L V F (7) L 1 1 L I V Γ V. (8) 1 Noie ha in (7) F is he imension-eni fis-oe suffiien saisis obaine in a o-sep poess. In he fis sep he inpu saisis F is hiene an enalize fo eah Gaussian suh ha F Σ 12 F μ hee μ an Σ ae he mean veo an ovaiane maix of he -h mixue of he UBM. In he seon sep F is fome base on he hiene an enalize saisis F in a simila manne as m in (4). he seon sep oespons o he sap fom mixue-eni o imension-eni as iniae by he hange of subsip in Fig. 1 o in Fig. 2 (a he boom igh one of he seon eangula box). Sine eah iniviual imension of he aousi feaue has is on laen vaiable a speeh ueane oul heeby be epesene by a se of poseio mean veos y 1 y2 y D D hih e efee o as he loal vaiabiliy veos in a simila vein as he i-veo exaion. he expeaion maximizaion (EM) algoihm is use o esimae he loaing maies V fo = 1 2 D. he loaing maies V oul be obaine by ieaively maximizing he folloing auxiliay funion: 1 R D QE V F V ΓV. (9) aking he eivaive of (9) ih espe o V an seing he eivaive o zeo e obain he fomula fo esimaing V hih onsiues he M-sep of he EM ieaion: E E ΓV F. (10) Moe speifially he -h o of V an be upae as: 1 v Φ K (11) hee Φ F E an K E. In (11) he supesip in Φ enoes he -h o of Φ an is he oupany of session in he -h Gaussian as (6) PLDA fo loal vaiabiliy veos Simila o i-veos he loal vaiabiliy veos also onain boh speake an hannel vaiabiliy making i neessay fo some fom of hannel ompensaion o be applie [11]. his objeive is ahieve ih PLDA [12]. Reall ha hee ae D loal vaiabiliy veos fo eah given ueane. We fom a loal vaiabiliy supeveo by onaenaing iniviual elemens as y1 y2 y D. Compae o he GMM mean supeveo one avanage of

4 his loal vaiabiliy supeveo is ha i has a muh loe imensionaliy hih allos he PLDA o be applie iely. I is oh menioning ha he imension euion is ahieve pe imension aoss all mixues. he ask of speake veifiaion is o eemine hehe o no an enollmen ueane an a es ueane ae fom he same speake [13 14]. In his pape e fisly obain an enollmen speake PLDA moel hough aapaion on he univesal moel [15]. hen he log-likelihoo aio of he es ueane beeen he enollmen an he univesal PLDA moels is ompue as: p e p l e log. (12) Hee he subsips e an enoe he enollmen an es sessions espeively. Deaile seps o evaluae he likelihoo an be foun in [15]. 4. Expeimens Expeimens ee aie ou on he elephone ials of he sho2-sho3 ask of NIS SRE 08 an he oe-oe ask of SRE 10. he nominal uaion of he aining an es segmens as abou o an a half minues. he pefomane as evaluae base on he equal-eo-ae (EER) an he minimum eeion os funion (mindcf). We onsie he mindcf a o iffeen opeaion poins namely mindcf08 an mindcf10. he aousi feaues ee 57-imensional veos of mel fequeny epsal oeffiiens (MFCC) ih fis an seon eivaives appene. We aine gene-epenen UBMs of 512 Gaussians ih full ovaiane maies using NIS SRE 04 aase. Fo i-veo e aine he oal vaiabiliy maix ih a ank of J = 400 using he elephone aa fom NIS SRE an 06. As suh he i-veo ha a imensionaliy of 400. he same aase as use o ain he PLDA moel ih an eigenvoie maix of ank 200 an eigenhannel maix of ank 50 an a full ovaiane maix. Fo he loal vaiabiliy moel D = 57 loaing maies V ih a ank of J = 20 ee aine using he same aase. his onfiguaion esule in 57 loal vaiabiliy veos of 20 imensions i.e. a loal vaiabiliy supeveo ih a imensionaliy of l140 is fome fo eah inpu ueane. he loal supeveos ee hen moele ih a PLDA moel ih F of ank 400 G of ank 300 an a iagonal esiual maix. he same aase as use fo he aining he PLDA fo boh ases. able I an able II ompae he pefomanes of VM an LVM on NIS SRE 08 an NIS SRE 10 espeively. Also shon in he ables ae he fusion esuls by a simple summaion of hei soes. VM folloe by PLDA (i.e. an i- veo PLDA sysem) is use as he baseline. he esuls onfim ha he loal vaiabiliy veos exae ih LVM ae effeive fo speake haaeizaion even hough hee is sill a mino gap ompae o he baseline i-veo PLDA. he fusion sysem oupefoms he baseline i-veo sysem in mos ases espeially in male ials shoing ha he loal able I: Pefomane ompaison of oal vaiabiliy moel (VM) loal vaiabiliy moel (LVM) an soe fusion on DE6 of sho2-sho3 ask in NIS SRE 08. Male VM LVM fusion Female VM LVM fusion able II: Pefomane ompaison of oal vaiabiliy moel (VM) loal vaiabiliy moel (LVM) an soe fusion on CC5 of oe-oe ess in NIS SRE 10. Male VM LVM fusion Female VM LVM fusion vaiabiliy veos aiional infomaion omplemenay o ha of he i-veo. his suggess ha he popose LVM moel he speake an session vaiabiliy ha is absen in he VM. 5. Conlusion We popose he loal vaiabiliy moel (LVM) fo exaing he loal vaiabiliy assoiae ih eah imension of he aousi feaues fo ex-inepenen speake veifiaion. he majo iffeene beeen he LVM an oal vaiabiliy moel (VM) lies a hehe o no o ie he mixues aoss iffeen imensions of he aousi feaues oniione on some laen vaiables. his iffeene is illusae ih he use of pobabilisi gaphial moel in he pape. We also eive he poseio infeene an he EM seps fo paamee leaning. Expeimenal esuls onfim he effiay of he loal vaiabiliy veos in haaeizing he voal popey of a speake. he esuls also sugges ha he popose LVM moel he vaiabiliy ha is absen in he VM. 7. Aknolegemens his ok of Liping Chen as paially suppoe by he Naional Naue Siene Founaion of China (Gan No ) an he eleoni infomaion inusy evelopmen fun of China (Gan No ).

5 8. Refeenes [1] D.A. Reynols.F. Quaiei an R.B. Dumn Speake veifiaion using aape Gaussian mixue moel Digial Signal Poessing vol. 10 no. 1-3 pp [2]. Kinnunen an H. Li An ovevie of ex-inepenen speake eogniion: fom feaues o supeveos Speeh Communiaion vol. 52 no. 1 pp Jan [3] P. Kenny G. Boulianne P. Ouelle an P. Dumouhel Speake an session vaiabiliy in GMM-Base speake veifiaion IEEE ans. Auio Speeh an Language Poessing vol. 15 no. 4 pp May [4] N. Dehak P. Kenny R. Dehak P. Dumouhel an P. Ouelle Fon-en fao analysis fo speake veifiaion IEEE ans. Auio Speeh an Language Poessing vol. 19 no. 4 pp May [5] A. Hah S. Kajaeka an A. Solke Wihin-lass ovaiane nomalizaion fo SVM-base speake eogniion in Inenaional Confeene on Spoken Language Poessing Pisbugh PA USA Sepembe [6] C. M. Bishop Paen Reogniion an Mahine Leaning. Spinge [7] S. J. D. Pine an J. H. Ele Pobabilisi linea isiminan analysis fo infeenes abou ieniy in Po. Inenaional Confeene on Compue Vision [8] P. Kenny. Safylakis P. Ouelle M. J. Alam an P. Dumouhel PLDA fo speake veifiaion ih ueane of abiay uaion in Po. IEEE ICASSP 2013 pp [9] P. Kenny A Small Foopin i-veo Exao in Po. Oyssey: Speake an Language Reogniion Wokshop June [10] P. Maejka O. Glembek F. Casalo J. Alam O. Plho P. Kenny L. Buge an J. Cenoky Full-ovaiane ubm an heavy-aile pla in i-veo speake veifiaion in Po. IEEE ICASSP 2011 pp [11] P. Kenny Bayesian speake veifiaion ih heavy-aile pios in Po. Oyssey: Speake an Language Reogniion Wokshop Jun [12] S. J. D. Pine Compue vision: moels leaning an infeene. Cambige Univesiy Pess [13] Y. Jiang K. A. Lee Z. ang B. Ma A. Lahe an H. Li PLDA moeling in i-veo an supeveo spae fo speake veifiaion in Po. INERSPEECH 2012 pape 198. [14] K. A. Lee A. Lahe C. H. You B. Ma H. Li Muli-session PLDA soing of i-veo fo paially open-se speake eeion in Po. INERSPEECH 2013 pp [15] L. Chen K. A. Lee B. Ma W. Guo H. Li an L. R. Dai Minimum ivegene esimaion of speake pio in muli-session PLDA soing in Po. ICASSP 2014 pp [16] N. Bümme an J. u Peez Appliaion-inepenen evaluaion of speake eeion Compue Speeh & Language vol. 20 no. 2 pp

LOCAL VARIABILITY MODELING FOR TEXT-INDEPENDENT SPEAKER VERIFICATION

LOCAL VARIABILITY MODELING FOR TEXT-INDEPENDENT SPEAKER VERIFICATION Odyssey 2014: he Speake and Language eognition Wokshop 16-19 June 2014, Joensuu, Finland LOAL VAIABILIY MODELING FO EX-INDEPENDEN SPEAKE VEIFIAION Liping hen 1, Kong Aik Lee 2, Bin Ma 2, Wu Guo 1, Haizhou

More information

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

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

More information

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

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

More information

Computer Propagation Analysis Tools

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

More information

STUDY OF THE STRESS-STRENGTH RELIABILITY AMONG THE PARAMETERS OF GENERALIZED INVERSE WEIBULL DISTRIBUTION

STUDY OF THE STRESS-STRENGTH RELIABILITY AMONG THE PARAMETERS OF GENERALIZED INVERSE WEIBULL DISTRIBUTION Inenaional Jounal of Science, Technology & Managemen Volume No 04, Special Issue No. 0, Mach 205 ISSN (online): 2394-537 STUDY OF THE STRESS-STRENGTH RELIABILITY AMONG THE PARAMETERS OF GENERALIZED INVERSE

More information

Coupled Mass Transport and Reaction in LPCVD Reactors

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

More information

An Automatic Door Sensor Using Image Processing

An Automatic Door Sensor Using Image Processing An Auomaic Doo Senso Using Image Pocessing Depamen o Elecical and Eleconic Engineeing Faculy o Engineeing Tooi Univesiy MENDEL 2004 -Insiue o Auomaion and Compue Science- in BRNO CZECH REPUBLIC 1. Inoducion

More information

Lecture 9. Transport Properties in Mesoscopic Systems. Over the last 1-2 decades, various techniques have been developed to synthesize

Lecture 9. Transport Properties in Mesoscopic Systems. Over the last 1-2 decades, various techniques have been developed to synthesize eue 9. anspo Popeies in Mesosopi Sysems Ove he las - deades, vaious ehniques have been developed o synhesize nanosuued maeials and o fabiae nanosale devies ha exhibi popeies midway beween he puely quanum

More information

Control Volume Derivation

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

More information

Real-Time Hand Tracking and Gesture Recognition System

Real-Time Hand Tracking and Gesture Recognition System GVIP 5 Confeene, 9- Deembe 5, CICC, Caio, Egp Real-ime Hand aing and Gesue Reogniion Ssem Nguen Dang Binh, Enoida Shuihi, oshiai Ejima Inelligene Media Laboao, Kushu Insiue of ehnolog 68-4, Kawazu, Iizua,

More information

Impact of I/O and Execution Scheduling Strategies on Large Scale Parallel Data Mining Nunnapus Benjamas, Putchong Uthayopas

Impact of I/O and Execution Scheduling Strategies on Large Scale Parallel Data Mining Nunnapus Benjamas, Putchong Uthayopas Impa of I/O and Exeuion Sheduling Saegies on Lage Sale Paallel Daa Mining... 648 unnapus enamas, Puhong Uhaopas Chaaeisi Evaluaion fo Goups in Daa Envelopmen nalsis and is ppliaion... 655 Kana KUROZUMI,

More information

Final Exam. Tuesday, December hours, 30 minutes

Final Exam. Tuesday, December hours, 30 minutes an Faniso ae Univesi Mihael Ba ECON 30 Fall 04 Final Exam Tuesda, Deembe 6 hous, 30 minues Name: Insuions. This is losed book, losed noes exam.. No alulaos of an kind ae allowed. 3. how all he alulaions.

More information

Support Vector Machines

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

More information

A Resonant Switched Reluctance Motor Drive for Marine Propulsion

A Resonant Switched Reluctance Motor Drive for Marine Propulsion A Resonan Swihed Reluane Moo Dive fo Maine Populsion Y.P.B.YEUNG, K.W.E.CHENG, S..HO and X.D.XUE Depamen of Eleial Engineeing The Hong Kong Polyehni Univesiy Hung Hom, Hong Kong SAR CHINA Absa: - Swihed

More information

Reinforcement learning

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

More information

CS 188: Artificial Intelligence Fall Probabilistic Models

CS 188: Artificial Intelligence Fall Probabilistic Models CS 188: Aificial Inelligence Fall 2007 Lecue 15: Bayes Nes 10/18/2007 Dan Klein UC Bekeley Pobabilisic Models A pobabilisic model is a join disibuion ove a se of vaiables Given a join disibuion, we can

More information

Time-Space Model of Business Fluctuations

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

More information

On The Estimation of Two Missing Values in Randomized Complete Block Designs

On The Estimation of Two Missing Values in Randomized Complete Block Designs Mahemaical Theoy and Modeling ISSN 45804 (Pape ISSN 505 (Online Vol.6, No.7, 06 www.iise.og On The Esimaion of Two Missing Values in Randomized Complee Bloc Designs EFFANGA, EFFANGA OKON AND BASSE, E.

More information

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

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

More information

Low-complexity Algorithms for MIMO Multiplexing Systems

Low-complexity Algorithms for MIMO Multiplexing Systems Low-complexiy Algoihms fo MIMO Muliplexing Sysems Ouline Inoducion QRD-M M algoihm Algoihm I: : o educe he numbe of suviving pahs. Algoihm II: : o educe he numbe of candidaes fo each ansmied signal. :

More information

[ ] 0. = (2) = a q dimensional vector of observable instrumental variables that are in the information set m constituents of u

[ ] 0. = (2) = a q dimensional vector of observable instrumental variables that are in the information set m constituents of u Genealized Mehods of Momens he genealized mehod momens (GMM) appoach of Hansen (98) can be hough of a geneal pocedue fo esing economics and financial models. he GMM is especially appopiae fo models ha

More information

Int. J. Computers & Electrical Engineering, vol. 30, no. 1, pp , AN OBSERVER DESIGN PROCEDURE FOR A CLASS OF NONLINEAR TIME-DELAY SYSTEMS

Int. J. Computers & Electrical Engineering, vol. 30, no. 1, pp , AN OBSERVER DESIGN PROCEDURE FOR A CLASS OF NONLINEAR TIME-DELAY SYSTEMS In. J. Compues & Elecical Engineeing, vol. 3, no., pp. 6-7, 3. AN OBSERVER DESIGN PROCEDURE FOR A CLASS OF NONLINEAR TIME-DELAY SYSTEMS * ** H. Tinh *, M. Aleen ** an S. Nahavani * School o Engineeing

More information

Special Subject SC434L Digital Video Coding and Compression. ASSIGNMENT 1-Solutions Due Date: Friday 30 August 2002

Special Subject SC434L Digital Video Coding and Compression. ASSIGNMENT 1-Solutions Due Date: Friday 30 August 2002 SC434L DVCC Assignmen Special Sujec SC434L Digial Vieo Coing an Compession ASSINMENT -Soluions Due Dae: Fiay 30 Augus 2002 This assignmen consiss of wo pages incluing wo compulsoy quesions woh of 0% of

More information

Projection of geometric models

Projection of geometric models ojecion of geomeic moels Copigh@, YZU Opimal Design Laboao. All ighs eseve. Las upae: Yeh-Liang Hsu (-9-). Noe: his is he couse maeial fo ME55 Geomeic moeling an compue gaphics, Yuan Ze Univesi. a of his

More information

Frequency-domain Eigenbeam-SDM and Equalization for High Speed Data Transmissions

Frequency-domain Eigenbeam-SDM and Equalization for High Speed Data Transmissions Fequeny-domain Eigenbeam-SDM and Equaizaion fo igh Speed Daa ansmissions Kazuyui Ozai Ainoi aajima and Fumiyui Adahi Dep. of Eeia and ommuniaions Engineeing Gaduaed shoo of Engineeing ohou Univesiy Sendai

More information

Millennium Theory Equations Original Copyright 2002 Joseph A. Rybczyk Updated Copyright 2003 Joseph A. Rybczyk Updated March 16, 2006

Millennium Theory Equations Original Copyright 2002 Joseph A. Rybczyk Updated Copyright 2003 Joseph A. Rybczyk Updated March 16, 2006 Millennim heoy Eqaions Oiginal Copyigh 00 Joseph A. Rybzyk Updaed Copyigh 003 Joseph A. Rybzyk Updaed Mah 6, 006 Following is a omplee lis o he Millennim heoy o Relaiviy eqaions: Fo easy eeene, all eqaions

More information

Part I. Labor- Leisure Decision (15 pts)

Part I. Labor- Leisure Decision (15 pts) Eon 509 Sping 204 Final Exam S. Paene Pa I. Labo- Leisue Deision (5 ps. Conside he following sai eonom given b he following equaions. Uili ln( H ln( l whee H sands fo he househ f f Poduion: Ah whee f sands

More information

Engineering Accreditation. Heat Transfer Basics. Assessment Results II. Assessment Results. Review Definitions. Outline

Engineering Accreditation. Heat Transfer Basics. Assessment Results II. Assessment Results. Review Definitions. Outline Hea ansfe asis Febua 7, 7 Hea ansfe asis a Caeo Mehanial Engineeing 375 Hea ansfe Febua 7, 7 Engineeing ediaion CSUN has aedied pogams in Civil, Eleial, Manufauing and Mehanial Engineeing Naional aediing

More information

Today - Lecture 13. Today s lecture continue with rotations, torque, Note that chapters 11, 12, 13 all involve rotations

Today - Lecture 13. Today s lecture continue with rotations, torque, Note that chapters 11, 12, 13 all involve rotations Today - Lecue 13 Today s lecue coninue wih oaions, oque, Noe ha chapes 11, 1, 13 all inole oaions slide 1 eiew Roaions Chapes 11 & 1 Viewed fom aboe (+z) Roaional, o angula elociy, gies angenial elociy

More information

Add the transfer payments and subtract the total taxes from (1): Using these definitions, the above becomes: The Saving and Investment Equation:

Add the transfer payments and subtract the total taxes from (1): Using these definitions, the above becomes: The Saving and Investment Equation: San Fanio Sae Univei Mihael Ba ECON 30 Fall 06 Poblem e 5 Conumpion-Saving Deiion and Riadian Equivalene. (0 poin. Saving and Invemen equaion. a. Deive he aving and invemen equaion. The fomula fo GDP uing

More information

The Transducer Influence on the Detection of a Transient Ultrasonic Field Scattered by a Rigid Point Target

The Transducer Influence on the Detection of a Transient Ultrasonic Field Scattered by a Rigid Point Target Rev. Eneg. Ren. : Physique Enegéique 1998) 49-56 The Tansdue Influene on he Deeion of a Tansien Ulasoni Field aeed by a Rigid Poin Tage H. Khelladi * and H. Djelouah ** * Insiu d Eleonique, ** Insiu de

More information

Lecture 22 Electromagnetic Waves

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

More information

Classification of Equations Characteristics

Classification of Equations Characteristics Clssiiion o Eqions Cheisis Consie n elemen o li moing in wo imensionl spe enoe s poin P elow. The ph o P is inie he line. The posiion ile is s so h n inemenl isne long is s. Le he goening eqions e epesene

More information

1 Temperature And Super Conductivity. 1.1 Defining Temperature

1 Temperature And Super Conductivity. 1.1 Defining Temperature 1 Tempeaue And Supe Conduiviy 1.1 Defining Tempeaue In ode o fully undesand his wok on empeaue and he elaed effes i helps o have ead he Quanum Theoy and he Advaned Quanum Theoy piees of he Pi-Spae Theoy

More information

Linear Quadratic Regulator (LQR) - State Feedback Design

Linear Quadratic Regulator (LQR) - State Feedback Design Linear Quadrai Regulaor (LQR) - Sae Feedbak Design A sysem is expressed in sae variable form as x = Ax + Bu n m wih x( ) R, u( ) R and he iniial ondiion x() = x A he sabilizaion problem using sae variable

More information

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

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

More information

Two-dimensional Effects on the CSR Interaction Forces for an Energy-Chirped Bunch. Rui Li, J. Bisognano, R. Legg, and R. Bosch

Two-dimensional Effects on the CSR Interaction Forces for an Energy-Chirped Bunch. Rui Li, J. Bisognano, R. Legg, and R. Bosch Two-dimensional Effecs on he CS Ineacion Foces fo an Enegy-Chiped Bunch ui Li, J. Bisognano,. Legg, and. Bosch Ouline 1. Inoducion 2. Pevious 1D and 2D esuls fo Effecive CS Foce 3. Bunch Disibuion Vaiaion

More information

Fall 2014 Final Exam (250 pts)

Fall 2014 Final Exam (250 pts) Eon 509 Fall 04 Final Exam (50 ps S. Paene Pa I. Answe ONLY ONE of he quesions below. (45 poins. Explain he onep of Riadian Equivalene. Wha ae he ondiions ha mus be saisfied fo i o h? Riadian Equivalene-

More information

Chapter Three Systems of Linear Differential Equations

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

More information

SKP-2 ALGORITHM: ON FORMING PART AND MACHINE CLUSTERS SEPARATELY

SKP-2 ALGORITHM: ON FORMING PART AND MACHINE CLUSTERS SEPARATELY Poeedings of the 1998 Paifi Confeene on Manufatuing, August 18-20, 1998, Bisbane, Queensland, Austalia SKP-2 ALGORITHM: ON FORMING PART AND MACHINE CLUSTERS SEPARATELY Susanto,S., Kennedy,R.D. and Pie,

More information

CROSSTALK ANALYSIS FOR HIGH-PRECISION OPTICAL PICKUP ACTUATOR SYSTEM

CROSSTALK ANALYSIS FOR HIGH-PRECISION OPTICAL PICKUP ACTUATOR SYSTEM Jounal o Theoeial and Applied Inomaion Tehnology h Apil 2. Vol. 5 No. 25-2 JATIT & LLS. All ighs eseved. ISSN: 992-8645 www.jai.og E-ISSN: 87-95 CROSSTALK ANALYSIS FOR HIGH-PRECISION OPTICAL PICKUP ACTUATOR

More information

Kalman Filter: an instance of Bayes Filter. Kalman Filter: an instance of Bayes Filter. Kalman Filter. Linear dynamics with Gaussian noise

Kalman Filter: an instance of Bayes Filter. Kalman Filter: an instance of Bayes Filter. Kalman Filter. Linear dynamics with Gaussian noise COM47 Inoducion o Roboics and Inelligen ysems he alman File alman File: an insance of Bayes File alman File: an insance of Bayes File Linea dynamics wih Gaussian noise alman File Linea dynamics wih Gaussian

More information

0.1 MAXIMUM LIKELIHOOD ESTIMATION EXPLAINED

0.1 MAXIMUM LIKELIHOOD ESTIMATION EXPLAINED 0.1 MAXIMUM LIKELIHOOD ESTIMATIO EXPLAIED Maximum likelihood esimaion is a bes-fi saisical mehod for he esimaion of he values of he parameers of a sysem, based on a se of observaions of a random variable

More information

Lecture 18: Kinetics of Phase Growth in a Two-component System: general kinetics analysis based on the dilute-solution approximation

Lecture 18: Kinetics of Phase Growth in a Two-component System: general kinetics analysis based on the dilute-solution approximation Lecue 8: Kineics of Phase Gowh in a Two-componen Sysem: geneal kineics analysis based on he dilue-soluion appoximaion Today s opics: In he las Lecues, we leaned hee diffeen ways o descibe he diffusion

More information

Nevertheless, there are well defined (and potentially useful) distributions for which σ 2

Nevertheless, there are well defined (and potentially useful) distributions for which σ 2 M. Meseron-Gibbons: Bioalulus, Leure, Page. The variane. More on improper inegrals In general, knowing only he mean of a isribuion is no as useful as also knowing wheher he isribuion is lumpe near he mean

More information

Sections 3.1 and 3.4 Exponential Functions (Growth and Decay)

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

More information

Projection of geometric models

Projection of geometric models ojecion of geomeic moels Eie: Yeh-Liang Hsu (998-9-2); ecommene: Yeh-Liang Hsu (2-9-26); las upae: Yeh-Liang Hsu (29--3). Noe: This is he couse maeial fo ME55 Geomeic moeling an compue gaphics, Yuan Ze

More information

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

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

More information

FOUR-WHEEL VEHICLE SUSPENSION MODELING FOR CONTROL SYSTEM DEVELOPMENT

FOUR-WHEEL VEHICLE SUSPENSION MODELING FOR CONTROL SYSTEM DEVELOPMENT Poeedings of COEM 5 Copyigh 5 by CM 8h Inenaional Congess of Mehanial Engineeing Novembe 6-, 5, Ouo Peo, MG FOUR-WHEEL VEHICLE SUSPENSION MODELING FOR CONROL SYSEM DEVELOPMEN Cláudio Civellao DN Suual

More information

Online Completion of Ill-conditioned Low-Rank Matrices

Online Completion of Ill-conditioned Low-Rank Matrices Online Compleion of Ill-condiioned Low-Rank Maices Ryan Kennedy and Camillo J. Taylo Compue and Infomaion Science Univesiy of Pennsylvania Philadelphia, PA, USA keny, cjaylo}@cis.upenn.edu Laua Balzano

More information

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

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

More information

Probabilistic Models. CS 188: Artificial Intelligence Fall Independence. Example: Independence. Example: Independence? Conditional Independence

Probabilistic Models. CS 188: Artificial Intelligence Fall Independence. Example: Independence. Example: Independence? Conditional Independence C 188: Aificial Inelligence Fall 2007 obabilisic Models A pobabilisic model is a join disibuion ove a se of vaiables Lecue 15: Bayes Nes 10/18/2007 Given a join disibuion, we can eason abou unobseved vaiables

More information

D.I. Survival models and copulas

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

More information

Combinatorial Approach to M/M/1 Queues. Using Hypergeometric Functions

Combinatorial Approach to M/M/1 Queues. Using Hypergeometric Functions Inenaional Mahemaical Foum, Vol 8, 03, no 0, 463-47 HIKARI Ld, wwwm-hikaicom Combinaoial Appoach o M/M/ Queues Using Hypegeomeic Funcions Jagdish Saan and Kamal Nain Depamen of Saisics, Univesiy of Delhi,

More information

Photographing a time interval

Photographing a time interval Potogaping a time inteval Benad Rotenstein and Ioan Damian Politennia Univesity of imisoaa Depatment of Pysis imisoaa Romania benad_otenstein@yaoo.om ijdamian@yaoo.om Abstat A metod of measuing time intevals

More information

Sharif University of Technology - CEDRA By: Professor Ali Meghdari

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

More information

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

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

More information

4/3/2017. PHY 712 Electrodynamics 9-9:50 AM MWF Olin 103

4/3/2017. PHY 712 Electrodynamics 9-9:50 AM MWF Olin 103 PHY 7 Eleodnais 9-9:50 AM MWF Olin 0 Plan fo Leue 0: Coninue eading Chap Snhoon adiaion adiaion fo eleon snhoon deies adiaion fo asonoial objes in iula obis 0/05/07 PHY 7 Sping 07 -- Leue 0 0/05/07 PHY

More information

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

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

More information

Orthotropic Materials

Orthotropic Materials Kapiel 2 Ohoopic Maeials 2. Elasic Sain maix Elasic sains ae elaed o sesses by Hooke's law, as saed below. The sesssain elaionship is in each maeial poin fomulaed in he local caesian coodinae sysem. ε

More information

Quantum Algorithms for Matrix Products over Semirings

Quantum Algorithms for Matrix Products over Semirings CHICAGO JOURNAL OF THEORETICAL COMPUTER SCIENCE 2017, Aicle 1, pages 1 25 hp://cjcscsuchicagoedu/ Quanum Algoihms fo Maix Poducs ove Semiings Fançois Le Gall Haumichi Nishimua Received July 24, 2015; Revised

More information

Velocity and Acceleration Simulation of a Vehicle with a Continuously Variable Power Split Transmission

Velocity and Acceleration Simulation of a Vehicle with a Continuously Variable Power Split Transmission Wold Aademy of Siene, Engineeing and Tehnology 55 009 eloiy and Aeleaion Simulaion of a ehile wih a Coninuously aiable Powe Spli Tansmission A. Babaei, N. Choupani Absa A oninuously vaiable ansmission

More information

CptS 570 Machine Learning School of EECS Washington State University. CptS Machine Learning 1

CptS 570 Machine Learning School of EECS Washington State University. CptS Machine Learning 1 ps 57 Machne Leann School of EES Washnon Sae Unves ps 57 - Machne Leann Assume nsances of classes ae lneal sepaable Esmae paamees of lnea dscmnan If ( - -) > hen + Else - ps 57 - Machne Leann lassfcaon

More information

and substitute in to the 1 st period budget constraint b. Derive the utility maximization condition (10 pts)

and substitute in to the 1 st period budget constraint b. Derive the utility maximization condition (10 pts) Eon 50 Sping 06 Midem Examinaion (00 ps S.L. Paene Ovelapping Geneaions Model (50 ps Conside he following he Ovelapping Geneaions model whee people live wo peiods. Eah geneaion has he same numbe of people.

More information

An analysis of precise positioning scenarios of the electromechanical rotating system driven by a stepping motor

An analysis of precise positioning scenarios of the electromechanical rotating system driven by a stepping motor SIRM h Inenaional Confeene on Vibaions in Roaing Mahines Magdebug Gemany.. Febuay An analysis of peise posiioning senaios of he eleomehanial oaing sysem diven by a sepping moo Robe Konowoi Andzej Pohane

More information

Physics 442. Electro-Magneto-Dynamics. M. Berrondo. Physics BYU

Physics 442. Electro-Magneto-Dynamics. M. Berrondo. Physics BYU Physis 44 Eleo-Magneo-Dynamis M. Beondo Physis BYU Paaveos Φ= V + Α Φ= V Α = = + J = + ρ J J ρ = J S = u + em S S = u em S Physis BYU Poenials Genealize E = V o he ime dependen E & B ase Podu of paaveos:

More information

Chapter Finite Difference Method for Ordinary Differential Equations

Chapter Finite Difference Method for Ordinary Differential Equations Chape 8.7 Finie Diffeence Mehod fo Odinay Diffeenial Eqaions Afe eading his chape, yo shold be able o. Undesand wha he finie diffeence mehod is and how o se i o solve poblems. Wha is he finie diffeence

More information

A PARAMETRIC REPRESENTATION OF RULED SURFACES

A PARAMETRIC REPRESENTATION OF RULED SURFACES A PARAMETRIC REPRESENTATION OF RULED SURFACES ELENA PROUSALIDOU, SEAN HANNA Univesiy College London, Unied Kingdom Asa. This pape poposes a simple paamei sysem o geneae an almos omplee se of uled sufaes

More information

Optimal Transform: The Karhunen-Loeve Transform (KLT)

Optimal Transform: The Karhunen-Loeve Transform (KLT) Opimal ransform: he Karhunen-Loeve ransform (KL) Reall: We are ineresed in uniary ransforms beause of heir nie properies: energy onservaion, energy ompaion, deorrelaion oivaion: τ (D ransform; assume separable)

More information

Review and Walkthrough of 1D FDTD

Review and Walkthrough of 1D FDTD 6//8 533 leomagnei Analsis Using Finie Diffeene Time Domain Leue #8 Review and Walhough of D FDTD Leue 8These noes ma onain opighed maeial obained unde fai use ules. Disibuion of hese maeials is sil pohibied

More information

EVENT HORIZONS IN COSMOLOGY

EVENT HORIZONS IN COSMOLOGY Mahemaics Today Vol7(Dec-)54-6 ISSN 976-38 EVENT HORIZONS IN COSMOLOGY K Punachanda Rao Depamen of Mahemaics Chiala Engineeing College Chiala 53 57 Andha Padesh, INDIA E-mail: dkpaocecc@yahoocoin ABSTRACT

More information

CYCLOSTATIONARITY-BASED BLIND CLASSIFICATION OF ANALOG AND DIGITAL MODULATIONS

CYCLOSTATIONARITY-BASED BLIND CLASSIFICATION OF ANALOG AND DIGITAL MODULATIONS CYCLOSAIONARIY-BASED BLIND CLASSIFICAION OF ANALOG AND DIGIAL MODULAIONS Oavia A. Dobe Ali Abdi 2 Yehekel Ba-Ne 2 Wei Su 3 Fauly of Eng. Applied Siene Memoial Univeiy of Newfoundl S. John NL AB 3X5 Canada

More information

EFFECT OF PERMISSIBLE DELAY ON TWO-WAREHOUSE INVENTORY MODEL FOR DETERIORATING ITEMS WITH SHORTAGES

EFFECT OF PERMISSIBLE DELAY ON TWO-WAREHOUSE INVENTORY MODEL FOR DETERIORATING ITEMS WITH SHORTAGES Volume, ssue 3, Mach 03 SSN 39-4847 EFFEC OF PERMSSBLE DELAY ON WO-WAREHOUSE NVENORY MODEL FOR DEERORANG EMS WH SHORAGES D. Ajay Singh Yadav, Ms. Anupam Swami Assisan Pofesso, Depamen of Mahemaics, SRM

More information

WORK POWER AND ENERGY Consevaive foce a) A foce is said o be consevaive if he wok done by i is independen of pah followed by he body b) Wok done by a consevaive foce fo a closed pah is zeo c) Wok done

More information

for Model Selection in the AR(1)

for Model Selection in the AR(1) Loal o Uniy, Long-Hoizon Foeasing hesholds fo Model Seleion in he AR John L. une # Absa: he pape develops a famewok fo analyzing long-hoizon foeasing in he AR model using he loal o uniy speifiaion of he

More information

Research on the Algorithm of Evaluating and Analyzing Stationary Operational Availability Based on Mission Requirement

Research on the Algorithm of Evaluating and Analyzing Stationary Operational Availability Based on Mission Requirement Reseach on he Algoihm of Evaluaing and Analyzing Saionay Opeaional Availabiliy Based on ission Requiemen Wang Naichao, Jia Zhiyu, Wang Yan, ao Yilan, Depamen of Sysem Engineeing of Engineeing Technology,

More information

Mass Transfer Coefficients (MTC) and Correlations I

Mass Transfer Coefficients (MTC) and Correlations I Mass Transfer Mass Transfer Coeffiiens (MTC) and Correlaions I 7- Mass Transfer Coeffiiens and Correlaions I Diffusion an be desribed in wo ways:. Deailed physial desripion based on Fik s laws and he diffusion

More information

Dynamic Estimation of OD Matrices for Freeways and Arterials

Dynamic Estimation of OD Matrices for Freeways and Arterials Novembe 2007 Final Repo: ITS Dynamic Esimaion of OD Maices fo Feeways and Aeials Auhos: Juan Calos Heea, Sauabh Amin, Alexande Bayen, Same Madana, Michael Zhang, Yu Nie, Zhen Qian, Yingyan Lou, Yafeng

More information

336 ERIDANI kfk Lp = sup jf(y) ; f () jj j p p whee he supemum is aken ove all open balls = (a ) inr n, jj is he Lebesgue measue of in R n, () =(), f

336 ERIDANI kfk Lp = sup jf(y) ; f () jj j p p whee he supemum is aken ove all open balls = (a ) inr n, jj is he Lebesgue measue of in R n, () =(), f TAMKANG JOURNAL OF MATHEMATIS Volume 33, Numbe 4, Wine 2002 ON THE OUNDEDNESS OF A GENERALIED FRATIONAL INTEGRAL ON GENERALIED MORREY SPAES ERIDANI Absac. In his pape we exend Nakai's esul on he boundedness

More information

Multicode DS-CDMA With Joint Transmit/Receive Frequency-domain Equalization

Multicode DS-CDMA With Joint Transmit/Receive Frequency-domain Equalization Muliode DS-CDMA Wih Join Tanmi/Reeive Fequeny-domain Equalizaion Kazuki TAKEDA iomihi TOMEBA and Fumiyuki ADACI Dep. of Eleial and Communiaion Engineeing, Gaduae Shool of Engineeing, Tohoku Univeiy 6-6-5,

More information

From E.G. Haug Escape Velocity To the Golden Ratio at the Black Hole. Branko Zivlak, Novi Sad, May 2018

From E.G. Haug Escape Velocity To the Golden Ratio at the Black Hole. Branko Zivlak, Novi Sad, May 2018 Fom E.G. Haug Esape eloity To the Golden Ratio at the Blak Hole Banko Zivlak, bzivlak@gmail.om Novi Sad, May 018 Abstat Esape veloity fom the E.G. Haug has been heked. It is ompaed with obital veloity

More information

PROCESS SIMULATING OF HEAT TRANSFER IN HIGH- TEMPERATURE THERMOCOUPLES

PROCESS SIMULATING OF HEAT TRANSFER IN HIGH- TEMPERATURE THERMOCOUPLES MAEC Web of Confeenes 0006 ( 05) DOI: 0.05/ maeonf/ 050006 C Owned by he auhos published by EDP Sienes 05 PROCESS SIMULAING OF HEA RANSFER IN HIGH- EMPERAURE HERMOCOUPLES Yuliana K. Aoshenko Alena A. Byhkova

More information

EN221 - Fall HW # 7 Solutions

EN221 - Fall HW # 7 Solutions EN221 - Fall2008 - HW # 7 Soluions Pof. Vivek Shenoy 1.) Show ha he fomulae φ v ( φ + φ L)v (1) u v ( u + u L)v (2) can be pu ino he alenaive foms φ φ v v + φv na (3) u u v v + u(v n)a (4) (a) Using v

More information

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

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

More information

Relative and Circular Motion

Relative and Circular Motion Relaie and Cicula Moion a) Relaie moion b) Cenipeal acceleaion Mechanics Lecue 3 Slide 1 Mechanics Lecue 3 Slide 2 Time on Video Pelecue Looks like mosly eeyone hee has iewed enie pelecue GOOD! Thank you

More information

Predictive Regressions. Based on AP Chap. 20

Predictive Regressions. Based on AP Chap. 20 Peicive Regessions Base on AP Chap. 20 Ealy auhos, incluing Jensen (969) an Fama (970) viewe ha he efficien mae hypohesis mean euns wee no peicable. Lae wo, noably Lucas (978) showe ha aional expecaions

More information

Lecture 17: Kinetics of Phase Growth in a Two-component System:

Lecture 17: Kinetics of Phase Growth in a Two-component System: Lecue 17: Kineics of Phase Gowh in a Two-componen Sysem: descipion of diffusion flux acoss he α/ ineface Today s opics Majo asks of oday s Lecue: how o deive he diffusion flux of aoms. Once an incipien

More information

Mathematical Foundations -1- Choice over Time. Choice over time. A. The model 2. B. Analysis of period 1 and period 2 3

Mathematical Foundations -1- Choice over Time. Choice over time. A. The model 2. B. Analysis of period 1 and period 2 3 Mahemaial Foundaions -- Choie over Time Choie over ime A. The model B. Analysis of period and period 3 C. Analysis of period and period + 6 D. The wealh equaion 0 E. The soluion for large T 5 F. Fuure

More information

Physics 2001/2051 Moments of Inertia Experiment 1

Physics 2001/2051 Moments of Inertia Experiment 1 Physics 001/051 Momens o Ineia Expeimen 1 Pelab 1 Read he ollowing backgound/seup and ensue you ae amilia wih he heoy equied o he expeimen. Please also ill in he missing equaions 5, 7 and 9. Backgound/Seup

More information

A Weighted Moving Average Process for Forecasting. Shou Hsing Shih Chris P. Tsokos

A Weighted Moving Average Process for Forecasting. Shou Hsing Shih Chris P. Tsokos A Weighed Moving Aveage Pocess fo Foecasing Shou Hsing Shih Chis P. Tsokos Depamen of Mahemaics and Saisics Univesiy of Souh Floida, USA Absac The objec of he pesen sudy is o popose a foecasing model fo

More information

mywbut.com Lesson 11 Study of DC transients in R-L-C Circuits

mywbut.com Lesson 11 Study of DC transients in R-L-C Circuits mywbu.om esson Sudy of DC ransiens in R--C Ciruis mywbu.om Objeives Be able o wrie differenial equaion for a d iruis onaining wo sorage elemens in presene of a resisane. To develop a horough undersanding

More information

MIMO Cognitive Radio Capacity in. Flat Fading Channel. Mohan Premkumar, Muthappa Perumal Chitra. 1. Introduction

MIMO Cognitive Radio Capacity in. Flat Fading Channel. Mohan Premkumar, Muthappa Perumal Chitra. 1. Introduction Inenaional Jounal of Wieless Communicaions, ewoking and Mobile Compuing 07; 4(6): 44-50 hp://www.aasci.og/jounal/wcnmc ISS: 38-37 (Pin); ISS: 38-45 (Online) MIMO Cogniive adio Capaciy in Fla Fading Channel

More information

KINGS UNIT- I LAPLACE TRANSFORMS

KINGS UNIT- I LAPLACE TRANSFORMS MA5-MATHEMATICS-II KINGS COLLEGE OF ENGINEERING Punalkulam DEPARTMENT OF MATHEMATICS ACADEMIC YEAR - ( Even Semese ) QUESTION BANK SUBJECT CODE: MA5 SUBJECT NAME: MATHEMATICS - II YEAR / SEM: I / II UNIT-

More information

Fig. 1S. The antenna construction: (a) main geometrical parameters, (b) the wire support pillar and (c) the console link between wire and coaxial

Fig. 1S. The antenna construction: (a) main geometrical parameters, (b) the wire support pillar and (c) the console link between wire and coaxial a b c Fig. S. The anenna consucion: (a) ain geoeical paaees, (b) he wie suppo pilla and (c) he console link beween wie and coaial pobe. Fig. S. The anenna coss-secion in he y-z plane. Accoding o [], he

More information

Overview. Overview Page 1 of 8

Overview. Overview Page 1 of 8 COMPUTERS AND STRUCTURES, INC., BERKELEY, CALIORNIA DECEMBER 2001 COMPOSITE BEAM DESIGN AISC-LRD93 Tecnical Noe Compac and Noncompac Requiemens Tis Tecnical Noe descibes o e pogam cecks e AISC-LRD93 specificaion

More information

CHAPTER 10: LINEAR DISCRIMINATION

CHAPTER 10: LINEAR DISCRIMINATION HAPER : LINEAR DISRIMINAION Dscmnan-based lassfcaon 3 In classfcaon h K classes ( k ) We defned dsmnan funcon g () = K hen gven an es eample e chose (pedced) s class label as f g () as he mamum among g

More information

The Production of Polarization

The Production of Polarization Physics 36: Waves Lecue 13 3/31/211 The Poducion of Polaizaion Today we will alk abou he poducion of polaized ligh. We aleady inoduced he concep of he polaizaion of ligh, a ansvese EM wave. To biefly eview

More information

Unsupervised Segmentation of Moving MPEG Blocks Based on Classification of Temporal Information

Unsupervised Segmentation of Moving MPEG Blocks Based on Classification of Temporal Information Unsupevised Segmenaion of Moving MPEG Blocs Based on Classificaion of Tempoal Infomaion Ofe Mille 1, Ami Avebuch 1, and Yosi Kelle 2 1 School of Compue Science,Tel-Aviv Univesiy, Tel-Aviv 69978, Isael

More information

KINEMATICS OF RIGID BODIES

KINEMATICS OF RIGID BODIES KINEMTICS OF RIGID ODIES In igid body kinemaics, we use he elaionships govening he displacemen, velociy and acceleaion, bu mus also accoun fo he oaional moion of he body. Descipion of he moion of igid

More information

Bayes Nets. CS 188: Artificial Intelligence Spring Example: Alarm Network. Building the (Entire) Joint

Bayes Nets. CS 188: Artificial Intelligence Spring Example: Alarm Network. Building the (Entire) Joint C 188: Aificial Inelligence ping 2008 Bayes Nes 2/5/08, 2/7/08 Dan Klein UC Bekeley Bayes Nes A Bayes ne is an efficien encoding of a pobabilisic model of a domain Quesions we can ask: Infeence: given

More information