Latent Spaces and Matrix Factorization

Size: px
Start display at page:

Download "Latent Spaces and Matrix Factorization"

Transcription

1 Compuaional Linguisics Laen Spaces and Marix Facorizaion Sefan Thaer & Dierich Klakow FR 4.7 Allgemeine Linguisik (Compuerlinguisik) Universiä des Saarlandes Summer 2013

2 Goal Goal: rea documen clusering and word clusering on he same fooing (same semanic space) find low dimensional represenaions

3 The word documen marix

4 Clusering Documen clusering words describe each documen by a vecor conaining he frequencies of he Word clusering describe each word by a vecor conaining he frequencies of is occurance in differen documen

5 Join word and documen clusering The word documen marix: Ener frequency (or f-idf) for each word and documen in a square scheme of numbers (marix)

6 Marices

7 Marices A marix is an array wih wo indices j=1...m e.g. in a pyhon program his could be A[i][j] wih i=1..n and Marix When wriing, ofen a subscrip noaion is used a... a A or a square scheme: a 1,1... N,1 a i, j... a 1, M... N, M a i, j Specific example of a 2x3 marix A

8 The ranspose of a marix The wo indices are swapped e.g. in a pyhon program his could be A[j][i]=A[i][j] for i=1..n and j=1...m Marix for he marices from he previous slide we have: A a a 1,1... M,1 a... j, i... a a 1, N... M, N Specific example of a 2x3 marix A Wha is A

9 Produc of wo marices The elemens of a produc marix can be calculaed in a pyhon program by for i in range(1,n+1): for j in range(1,m+1): Marix for k in range(1,k+1): C[i][j] = A[i][k]*B[k][j] In mah noaion C A B wih c i, j k K 1 a i, k b k, j Example see black board

10 Uni marix Uni marix: he elemen are he indicaor funcion a i, j i, j Example: Marix 1 A Ofen he uni marix is denoed by a 1

11 Orhogonal marices a marix A is orhogonal if 1 A A Marix Is he following marix orhogonal: A

12 Marices in pyhon See hp://wiki.scipy.org/tenaive_numpy_tuorial#head-a9063f71090f3d1fbbdae5397ccb4e882d2cf603

13

14 Laen Semanic Analysis (LSA) This secion mosly follows Manning and Schüze Chaper 15

15 Singular Value Decomposiion Decompose A such ha A ~ TSD Wih Marix ~ A A 2 minimal and T T 1 D D 1 A a by dmarix T a by n marix S a n by n marix D a dby n marix

16 An arificial Example of Singular Value Decomposiion Is T Marix S 2 2 D An SVD of 2 A

17 More realisic Example (from Manning and Schüze) Decompose Marix

18 More realisic Example (from Manning and Schüze) Marix

19 More realisic Example (from Manning and Schüze) Marix

20 Documen-Documen Similariy Rewrie A A d wih 1 Marix d j d 2... a vecor d wih word frequencies of d he j- h documen Similariy of i-h documen wih j-h documen d d i j All documen-documen similariies A A

21 Documen-Documen Similariy Rewrie Marix ~ A ~ A ( TSD DS DS ( SD T SD ) ) TSD TSD SD Measure similariy in subspace defined by SD

22 More realisic Example (from Manning and Schüze) Resul for SD Marix

23 More realisic Example (from Manning and Schüze) Decompose A such ha Marix

24 An even more realisic example Term Documen Marix Srucure

25 An even more realisic example Documen-Documen Similariy Term Documen Marix Srucure

26 Represenaion for Documens in 2 dimensional Subspace

27 Term-Term Similariy Rewrie Marix ~~ AA ( TSD TSD DS TS ( TS)( TS) ST )( TSD T ) Measure similariy in subspace defined by TS

28 Task How does your programming language suppor SVD Do some inerne search (~10 minues) Repor your findings Marix

29 Homework See shee Marix

30 LSA Performance LSA consisenly improves recall on sandard es collecions (precision/recall generally improved) Variable performance on larger TREC collecions Dimensionaliy of Laen Space a magic number seems o work fine no saisfacory way of assessing value. Compuaional cos high

31 Applicaion (by Landauer e. Al) Rae essay by similariy o exising ones Measure correlaion wih human raing Conclusion: drop he righ key-words and you are se

32 Probabilisic Laen Semanic Analysis (PLSA)

33 Moivaion Does orhogonally maer? Wouldn a sound saisical foundaion be beer?

34 PLSA Likelihood of documen P(doc) = P(erm 1 doc)p(erm 2 doc)... P(erm L doc) Inroduce erm-frequency marix X L l= 1 P(erm l doc)= T = 1 P(erm doc) A(erm,doc)

35 PLSA Inroduce hidden opic P(erm doc)= K k= 1 P(erm opic k )P(opic k doc) Shorhand =erm_ P( doc)= K k= 1 P( k)p(k doc) Relaion o LSA? Likelihood of documen P(doc) = T K = 1 k= 1 P( k)p(k doc) A(,doc)

36 PLSA: raining Training objecive funcion N d= 1 logp(d)= N T A(,d) log K d= 1 = 1 k= 1 P( k)p(k d) whichiso bemaximisedw. r.. parameersp( k) andhen alsop(k d), subjec o he consrains ha T = 1 P( k)= 1and K k= 1 P(k d)= 1.

37 PLSA: raining Updae erm-opic marix P1(,k) P1(,k) P1(,k) T = 1 N d= 1 P1(,k) P1(,k) K k= 1 A(,d) P1(,k)P2(k,d) P2(k,d) Updae opic-documen marix P2(k,d) P2(k,d) P2(k,d) K k= 1 T = 1 P2(k,d) P2(k,d) K k= 1 A(,d) P1(,k) P1(,k)P2(k,d)

38 PLSA P( k) for some opics

39 Comparison LSA and PLSA From Th. Hofmann, 2000

40 Non-negaive Marix Facorizaion See:

41 NMF: idea Find space ha separaes clusers beer

42 NMF: he model Decomposion of a non-negaive marix X in wo marices W and H boh non-negaive A=WH A: N x M daa marix W: N x R source marix H: R x M mixure marix

43 NMF: he model Deermine W and H such ha he produc WH is as close as possible o A W and H are bound o be non-negaive values Possible merics Kullback-Leibler-Divergenz Frobenius-Norm D A WH 1 2 A WH 2

44 NMF: raining Updae H W ab ab = H =W ab ab W W WH AH WHH A ab ab ab ab Relaion o updae From PLSA? In case he denominaor vanishes, add a small number

45 Homework Implemen NMF for he marix from he las lecure

46 Summary Ways o find laen semanic spaces: LSA PLSA NMF Similar facorizaions Differen arge funcions and consrains

Latent Spaces and Matrix Factorization

Latent Spaces and Matrix Factorization Compuaional Linguisics Laen Spaces and Marix Facorizaion Dierich Klakow FR 4.7 Allgemeine Linguisik (Compuerlinguisik) Universiä des Saarlandes Summer 0 Goal Goal: rea documen clusering and word clusering

More information

Retrieval Models. Boolean and Vector Space Retrieval Models. Common Preprocessing Steps. Boolean Model. Boolean Retrieval Model

Retrieval Models. Boolean and Vector Space Retrieval Models. Common Preprocessing Steps. Boolean Model. Boolean Retrieval Model 1 Boolean and Vecor Space Rerieval Models Many slides in his secion are adaped from Prof. Joydeep Ghosh (UT ECE) who in urn adaped hem from Prof. Dik Lee (Univ. of Science and Tech, Hong Kong) Rerieval

More information

Model Reduction for Dynamical Systems Lecture 6

Model Reduction for Dynamical Systems Lecture 6 Oo-von-Guericke Universiä Magdeburg Faculy of Mahemaics Summer erm 07 Model Reducion for Dynamical Sysems ecure 6 v eer enner and ihong Feng Max lanck Insiue for Dynamics of Complex echnical Sysems Compuaional

More information

10. State Space Methods

10. State Space Methods . Sae Space Mehods. Inroducion Sae space modelling was briefly inroduced in chaper. Here more coverage is provided of sae space mehods before some of heir uses in conrol sysem design are covered in he

More information

MANY FACET, COMMON LATENT TRAIT POLYTOMOUS IRT MODEL AND EM ALGORITHM. Dimitar Atanasov

MANY FACET, COMMON LATENT TRAIT POLYTOMOUS IRT MODEL AND EM ALGORITHM. Dimitar Atanasov Pliska Sud. Mah. Bulgar. 20 (2011), 5 12 STUDIA MATHEMATICA BULGARICA MANY FACET, COMMON LATENT TRAIT POLYTOMOUS IRT MODEL AND EM ALGORITHM Dimiar Aanasov There are many areas of assessmen where he level

More information

Complexity of Inference in Topic Models

Complexity of Inference in Topic Models Complexiy of Inference in Topic Models David Sonag, Daniel M. Roy Massachuses Insiue of Technology {dsonag,droy}@csail.mi.edu Absrac We consider he compuaional complexiy of finding he MAP assignmen of

More information

Probabilistic learning

Probabilistic learning Probabilisic learning Charles Elkan November 8, 2012 Imporan: These lecure noes are based closely on noes wrien by Lawrence Saul. Tex may be copied direcly from his noes, or paraphrased. Also, hese ypese

More information

CHAPTER 10 VALIDATION OF TEST WITH ARTIFICAL NEURAL NETWORK

CHAPTER 10 VALIDATION OF TEST WITH ARTIFICAL NEURAL NETWORK 175 CHAPTER 10 VALIDATION OF TEST WITH ARTIFICAL NEURAL NETWORK 10.1 INTRODUCTION Amongs he research work performed, he bes resuls of experimenal work are validaed wih Arificial Neural Nework. From he

More information

References are appeared in the last slide. Last update: (1393/08/19)

References are appeared in the last slide. Last update: (1393/08/19) SYSEM IDEIFICAIO Ali Karimpour Associae Professor Ferdowsi Universi of Mashhad References are appeared in he las slide. Las updae: 0..204 393/08/9 Lecure 5 lecure 5 Parameer Esimaion Mehods opics o be

More information

Refinement of Document Clustering by Using NMF *

Refinement of Document Clustering by Using NMF * Refinemen of Documen Clusering by Using NMF * Hiroyuki Shinnou and Minoru Sasaki Deparmen of Compuer and Informaion Sciences, Ibaraki Universiy, 4-12-1 Nakanarusawa, Hiachi, Ibaraki JAPAN 316-8511 {shinnou,

More information

Advanced time-series analysis (University of Lund, Economic History Department)

Advanced time-series analysis (University of Lund, Economic History Department) Advanced ime-series analysis (Universiy of Lund, Economic Hisory Deparmen) 30 Jan-3 February and 6-30 March 01 Lecure 9 Vecor Auoregression (VAR) echniques: moivaion and applicaions. Esimaion procedure.

More information

Linear Gaussian State Space Models

Linear Gaussian State Space Models Linear Gaussian Sae Space Models Srucural Time Series Models Level and Trend Models Basic Srucural Model (BSM Dynamic Linear Models Sae Space Model Represenaion Level, Trend, and Seasonal Models Time Varying

More information

Wavelet Methods for Time Series Analysis. What is a Wavelet? Part I: Introduction to Wavelets and Wavelet Transforms. sines & cosines are big waves

Wavelet Methods for Time Series Analysis. What is a Wavelet? Part I: Introduction to Wavelets and Wavelet Transforms. sines & cosines are big waves Wavele Mehods for Time Series Analysis Par I: Inroducion o Waveles and Wavele Transforms waveles are analysis ools for ime series and images as a subjec, waveles are relaively new (983 o presen) a synhesis

More information

Lecture 23: I. Data Dependence II. Dependence Testing: Formulation III. Dependence Testers IV. Loop Parallelization V.

Lecture 23: I. Data Dependence II. Dependence Testing: Formulation III. Dependence Testers IV. Loop Parallelization V. Lecure 23: Array Dependence Analysis & Parallelizaion I. Daa Dependence II. Dependence Tesing: Formulaion III. Dependence Tesers IV. Loop Parallelizaion V. Loop Inerchange [ALSU 11.6, 11.7.8] Phillip B.

More information

Logic in computer science

Logic in computer science Logic in compuer science Logic plays an imporan role in compuer science Logic is ofen called he calculus of compuer science Logic plays a similar role in compuer science o ha played by calculus in he physical

More information

PENALIZED LEAST SQUARES AND PENALIZED LIKELIHOOD

PENALIZED LEAST SQUARES AND PENALIZED LIKELIHOOD PENALIZED LEAST SQUARES AND PENALIZED LIKELIHOOD HAN XIAO 1. Penalized Leas Squares Lasso solves he following opimizaion problem, ˆβ lasso = arg max β R p+1 1 N y i β 0 N x ij β j β j (1.1) for some 0.

More information

Fusion and Inference from Multiple Data Sources in a Commensurate Space

Fusion and Inference from Multiple Data Sources in a Commensurate Space Fusion and Inference from Muliple Daa Sources in a Commensurae Space Zhiliang Ma 1, David. Marchee 2 and Carey E. Priebe 1 1 Applied Mahemaics & Saisics, ohns Hopkins Universiy, Balimore, MD, USA 2 Naval

More information

Chapter 3 Boundary Value Problem

Chapter 3 Boundary Value Problem Chaper 3 Boundary Value Problem A boundary value problem (BVP) is a problem, ypically an ODE or a PDE, which has values assigned on he physical boundary of he domain in which he problem is specified. Le

More information

1.6. Slopes of Tangents and Instantaneous Rate of Change

1.6. Slopes of Tangents and Instantaneous Rate of Change 1.6 Slopes of Tangens and Insananeous Rae of Change When you hi or kick a ball, he heigh, h, in meres, of he ball can be modelled by he equaion h() 4.9 2 v c. In his equaion, is he ime, in seconds; c represens

More information

How to Deal with Structural Breaks in Practical Cointegration Analysis

How to Deal with Structural Breaks in Practical Cointegration Analysis How o Deal wih Srucural Breaks in Pracical Coinegraion Analysis Roselyne Joyeux * School of Economic and Financial Sudies Macquarie Universiy December 00 ABSTRACT In his noe we consider he reamen of srucural

More information

HW6: MRI Imaging Pulse Sequences (7 Problems for 100 pts)

HW6: MRI Imaging Pulse Sequences (7 Problems for 100 pts) HW6: MRI Imaging Pulse Sequences (7 Problems for 100 ps) GOAL The overall goal of HW6 is o beer undersand pulse sequences for MRI image reconsrucion. OBJECTIVES 1) Design a spin echo pulse sequence o image

More information

Vehicle Arrival Models : Headway

Vehicle Arrival Models : Headway Chaper 12 Vehicle Arrival Models : Headway 12.1 Inroducion Modelling arrival of vehicle a secion of road is an imporan sep in raffic flow modelling. I has imporan applicaion in raffic flow simulaion where

More information

State-Space Models. Initialization, Estimation and Smoothing of the Kalman Filter

State-Space Models. Initialization, Estimation and Smoothing of the Kalman Filter Sae-Space Models Iniializaion, Esimaion and Smoohing of he Kalman Filer Iniializaion of he Kalman Filer The Kalman filer shows how o updae pas predicors and he corresponding predicion error variances when

More information

Independent component analysis for nonminimum phase systems using H filters

Independent component analysis for nonminimum phase systems using H filters Independen componen analysis for nonminimum phase sysems using H filers Shuichi Fukunaga, Kenji Fujimoo Deparmen of Mechanical Science and Engineering, Graduae Shool of Engineering, Nagoya Universiy, Furo-cho,

More information

Chapter 5. Heterocedastic Models. Introduction to time series (2008) 1

Chapter 5. Heterocedastic Models. Introduction to time series (2008) 1 Chaper 5 Heerocedasic Models Inroducion o ime series (2008) 1 Chaper 5. Conens. 5.1. The ARCH model. 5.2. The GARCH model. 5.3. The exponenial GARCH model. 5.4. The CHARMA model. 5.5. Random coefficien

More information

EKF SLAM vs. FastSLAM A Comparison

EKF SLAM vs. FastSLAM A Comparison vs. A Comparison Michael Calonder, Compuer Vision Lab Swiss Federal Insiue of Technology, Lausanne EPFL) michael.calonder@epfl.ch The wo algorihms are described wih a planar robo applicaion in mind. Generalizaion

More information

Empirical Process Theory

Empirical Process Theory Empirical Process heory 4.384 ime Series Analysis, Fall 27 Reciaion by Paul Schrimpf Supplemenary o lecures given by Anna Mikusheva Ocober 7, 28 Reciaion 7 Empirical Process heory Le x be a real-valued

More information

Dynamic models for largedimensional. Yields on U.S. Treasury securities (3 months to 10 years) y t

Dynamic models for largedimensional. Yields on U.S. Treasury securities (3 months to 10 years) y t Dynamic models for largedimensional vecor sysems A. Principal componens analysis Suppose we have a large number of variables observed a dae Goal: can we summarize mos of he feaures of he daa using jus

More information

An introduction to the theory of SDDP algorithm

An introduction to the theory of SDDP algorithm An inroducion o he heory of SDDP algorihm V. Leclère (ENPC) Augus 1, 2014 V. Leclère Inroducion o SDDP Augus 1, 2014 1 / 21 Inroducion Large scale sochasic problem are hard o solve. Two ways of aacking

More information

EECS 2602 Winter Laboratory 3 Fourier series, Fourier transform and Bode Plots in MATLAB

EECS 2602 Winter Laboratory 3 Fourier series, Fourier transform and Bode Plots in MATLAB EECS 6 Winer 7 Laboraory 3 Fourier series, Fourier ransform and Bode Plos in MATLAB Inroducion: The objecives of his lab are o use MATLAB:. To plo periodic signals wih Fourier series represenaion. To obain

More information

Appendix to Online l 1 -Dictionary Learning with Application to Novel Document Detection

Appendix to Online l 1 -Dictionary Learning with Application to Novel Document Detection Appendix o Online l -Dicionary Learning wih Applicaion o Novel Documen Deecion Shiva Prasad Kasiviswanahan Huahua Wang Arindam Banerjee Prem Melville A Background abou ADMM In his secion, we give a brief

More information

A Framework for Efficient Document Ranking Using Order and Non Order Based Fitness Function

A Framework for Efficient Document Ranking Using Order and Non Order Based Fitness Function A Framework for Efficien Documen Ranking Using Order and Non Order Based Finess Funcion Hazra Imran, Adii Sharan Absrac One cenral problem of informaion rerieval is o deermine he relevance of documens

More information

Appendix. Purdue University. From the SelectedWorks of David D Nolte. David D Nolte, Purdue University

Appendix. Purdue University. From the SelectedWorks of David D Nolte. David D Nolte, Purdue University Purdue Universiy From he SelecedWorks of David D Nole 2015 Appendix David D Nole, Purdue Universiy Available a: hp://works.bepress.com/ddnole/14/ APPENDIX APPENDIX: Appendix:... 1 A.1 The Complex Plane...

More information

Time series model fitting via Kalman smoothing and EM estimation in TimeModels.jl

Time series model fitting via Kalman smoothing and EM estimation in TimeModels.jl Time series model fiing via Kalman smoohing and EM esimaion in TimeModels.jl Gord Sephen Las updaed: January 206 Conens Inroducion 2. Moivaion and Acknowledgemens....................... 2.2 Noaion......................................

More information

Two Coupled Oscillators / Normal Modes

Two Coupled Oscillators / Normal Modes Lecure 3 Phys 3750 Two Coupled Oscillaors / Normal Modes Overview and Moivaion: Today we ake a small, bu significan, sep owards wave moion. We will no ye observe waves, bu his sep is imporan in is own

More information

23.2. Representing Periodic Functions by Fourier Series. Introduction. Prerequisites. Learning Outcomes

23.2. Representing Periodic Functions by Fourier Series. Introduction. Prerequisites. Learning Outcomes Represening Periodic Funcions by Fourier Series 3. Inroducion In his Secion we show how a periodic funcion can be expressed as a series of sines and cosines. We begin by obaining some sandard inegrals

More information

Chapter 2. First Order Scalar Equations

Chapter 2. First Order Scalar Equations Chaper. Firs Order Scalar Equaions We sar our sudy of differenial equaions in he same way he pioneers in his field did. We show paricular echniques o solve paricular ypes of firs order differenial equaions.

More information

ACE 562 Fall Lecture 4: Simple Linear Regression Model: Specification and Estimation. by Professor Scott H. Irwin

ACE 562 Fall Lecture 4: Simple Linear Regression Model: Specification and Estimation. by Professor Scott H. Irwin ACE 56 Fall 005 Lecure 4: Simple Linear Regression Model: Specificaion and Esimaion by Professor Sco H. Irwin Required Reading: Griffihs, Hill and Judge. "Simple Regression: Economic and Saisical Model

More information

Short Introduction to Fractional Calculus

Short Introduction to Fractional Calculus . Shor Inroducion o Fracional Calculus Mauro Bologna Deparameno de Física, Faculad de Ciencias Universidad de Tarapacá, Arica, Chile email: mbologna@ua.cl Absrac In he pas few years fracional calculus

More information

Exploiting Hierarchical Structures for Unsupervised Feature Selection

Exploiting Hierarchical Structures for Unsupervised Feature Selection Exploiing Hierarchical Srucures for Unsupervised Feaure Selecion Suhang Wang Yilin Wang Jiliang Tang Charu Aggarwal Huan Liu Suhas Ranganah Absrac Feaure selecion has been proven o be effecive and efficien

More information

Learning to Process Natural Language in Big Data Environment

Learning to Process Natural Language in Big Data Environment CCF ADL 2015 Nanchang Oc 11, 2015 Learning o Process Naural Language in Big Daa Environmen Hang Li Noah s Ark Lab Huawei Technologies Par 2: Useful Deep Learning Tools Powerful Deep Learning Tools (Unsupervised

More information

dy dx = xey (a) y(0) = 2 (b) y(1) = 2.5 SOLUTION: See next page

dy dx = xey (a) y(0) = 2 (b) y(1) = 2.5 SOLUTION: See next page Assignmen 1 MATH 2270 SOLUTION Please wrie ou complee soluions for each of he following 6 problems (one more will sill be added). You may, of course, consul wih your classmaes, he exbook or oher resources,

More information

Random Processes 1/24

Random Processes 1/24 Random Processes 1/24 Random Process Oher Names : Random Signal Sochasic Process A Random Process is an exension of he concep of a Random variable (RV) Simples View : A Random Process is a RV ha is a Funcion

More information

Section 4.4 Logarithmic Properties

Section 4.4 Logarithmic Properties Secion. Logarihmic Properies 5 Secion. Logarihmic Properies In he previous secion, we derived wo imporan properies of arihms, which allowed us o solve some asic eponenial and arihmic equaions. Properies

More information

Technical Report Doc ID: TR March-2013 (Last revision: 23-February-2016) On formulating quadratic functions in optimization models.

Technical Report Doc ID: TR March-2013 (Last revision: 23-February-2016) On formulating quadratic functions in optimization models. Technical Repor Doc ID: TR--203 06-March-203 (Las revision: 23-Februar-206) On formulaing quadraic funcions in opimizaion models. Auhor: Erling D. Andersen Convex quadraic consrains quie frequenl appear

More information

We just finished the Erdős-Stone Theorem, and ex(n, F ) (1 1/(χ(F ) 1)) ( n

We just finished the Erdős-Stone Theorem, and ex(n, F ) (1 1/(χ(F ) 1)) ( n Lecure 3 - Kövari-Sós-Turán Theorem Jacques Versraëe jacques@ucsd.edu We jus finished he Erdős-Sone Theorem, and ex(n, F ) ( /(χ(f ) )) ( n 2). So we have asympoics when χ(f ) 3 bu no when χ(f ) = 2 i.e.

More information

WEEK-3 Recitation PHYS 131. of the projectile s velocity remains constant throughout the motion, since the acceleration a x

WEEK-3 Recitation PHYS 131. of the projectile s velocity remains constant throughout the motion, since the acceleration a x WEEK-3 Reciaion PHYS 131 Ch. 3: FOC 1, 3, 4, 6, 14. Problems 9, 37, 41 & 71 and Ch. 4: FOC 1, 3, 5, 8. Problems 3, 5 & 16. Feb 8, 018 Ch. 3: FOC 1, 3, 4, 6, 14. 1. (a) The horizonal componen of he projecile

More information

Product Integration. Richard D. Gill. Mathematical Institute, University of Utrecht, Netherlands EURANDOM, Eindhoven, Netherlands August 9, 2001

Product Integration. Richard D. Gill. Mathematical Institute, University of Utrecht, Netherlands EURANDOM, Eindhoven, Netherlands August 9, 2001 Produc Inegraion Richard D. Gill Mahemaical Insiue, Universiy of Urech, Neherlands EURANDOM, Eindhoven, Neherlands Augus 9, 21 Absrac This is a brief survey of produc-inegraion for biosaisicians. 1 Produc-Inegraion

More information

Content-Based Shape Retrieval Using Different Shape Descriptors: A Comparative Study Dengsheng Zhang and Guojun Lu

Content-Based Shape Retrieval Using Different Shape Descriptors: A Comparative Study Dengsheng Zhang and Guojun Lu Conen-Based Shape Rerieval Using Differen Shape Descripors: A Comparaive Sudy Dengsheng Zhang and Guojun Lu Gippsland School of Compuing and Informaion Technology Monash Universiy Churchill, Vicoria 3842

More information

m = 41 members n = 27 (nonfounders), f = 14 (founders) 8 markers from chromosome 19

m = 41 members n = 27 (nonfounders), f = 14 (founders) 8 markers from chromosome 19 Sequenial Imporance Sampling (SIS) AKA Paricle Filering, Sequenial Impuaion (Kong, Liu, Wong, 994) For many problems, sampling direcly from he arge disribuion is difficul or impossible. One reason possible

More information

ODEs II, Lecture 1: Homogeneous Linear Systems - I. Mike Raugh 1. March 8, 2004

ODEs II, Lecture 1: Homogeneous Linear Systems - I. Mike Raugh 1. March 8, 2004 ODEs II, Lecure : Homogeneous Linear Sysems - I Mike Raugh March 8, 4 Inroducion. In he firs lecure we discussed a sysem of linear ODEs for modeling he excreion of lead from he human body, saw how o ransform

More information

4.1 Other Interpretations of Ridge Regression

4.1 Other Interpretations of Ridge Regression CHAPTER 4 FURTHER RIDGE THEORY 4. Oher Inerpreaions of Ridge Regression In his secion we will presen hree inerpreaions for he use of ridge regression. The firs one is analogous o Hoerl and Kennard reasoning

More information

THE DISCRETE WAVELET TRANSFORM

THE DISCRETE WAVELET TRANSFORM . 4 THE DISCRETE WAVELET TRANSFORM 4 1 Chaper 4: THE DISCRETE WAVELET TRANSFORM 4 2 4.1 INTRODUCTION TO DISCRETE WAVELET THEORY The bes way o inroduce waveles is hrough heir comparison o Fourier ransforms,

More information

GMM - Generalized Method of Moments

GMM - Generalized Method of Moments GMM - Generalized Mehod of Momens Conens GMM esimaion, shor inroducion 2 GMM inuiion: Maching momens 2 3 General overview of GMM esimaion. 3 3. Weighing marix...........................................

More information

Course Notes for EE227C (Spring 2018): Convex Optimization and Approximation

Course Notes for EE227C (Spring 2018): Convex Optimization and Approximation Course Noes for EE7C Spring 018: Convex Opimizaion and Approximaion Insrucor: Moriz Hard Email: hard+ee7c@berkeley.edu Graduae Insrucor: Max Simchowiz Email: msimchow+ee7c@berkeley.edu Ocober 15, 018 3

More information

4.1 - Logarithms and Their Properties

4.1 - Logarithms and Their Properties Chaper 4 Logarihmic Funcions 4.1 - Logarihms and Their Properies Wha is a Logarihm? We define he common logarihm funcion, simply he log funcion, wrien log 10 x log x, as follows: If x is a posiive number,

More information

Isolated-word speech recognition using hidden Markov models

Isolated-word speech recognition using hidden Markov models Isolaed-word speech recogniion using hidden Markov models Håkon Sandsmark December 18, 21 1 Inroducion Speech recogniion is a challenging problem on which much work has been done he las decades. Some of

More information

Vectorautoregressive Model and Cointegration Analysis. Time Series Analysis Dr. Sevtap Kestel 1

Vectorautoregressive Model and Cointegration Analysis. Time Series Analysis Dr. Sevtap Kestel 1 Vecorauoregressive Model and Coinegraion Analysis Par V Time Series Analysis Dr. Sevap Kesel 1 Vecorauoregression Vecor auoregression (VAR) is an economeric model used o capure he evoluion and he inerdependencies

More information

Georey E. Hinton. University oftoronto. Technical Report CRG-TR February 22, Abstract

Georey E. Hinton. University oftoronto.   Technical Report CRG-TR February 22, Abstract Parameer Esimaion for Linear Dynamical Sysems Zoubin Ghahramani Georey E. Hinon Deparmen of Compuer Science Universiy oftorono 6 King's College Road Torono, Canada M5S A4 Email: zoubin@cs.orono.edu Technical

More information

L07. KALMAN FILTERING FOR NON-LINEAR SYSTEMS. NA568 Mobile Robotics: Methods & Algorithms

L07. KALMAN FILTERING FOR NON-LINEAR SYSTEMS. NA568 Mobile Robotics: Methods & Algorithms L07. KALMAN FILTERING FOR NON-LINEAR SYSTEMS NA568 Mobile Roboics: Mehods & Algorihms Today s Topic Quick review on (Linear) Kalman Filer Kalman Filering for Non-Linear Sysems Exended Kalman Filer (EKF)

More information

Diebold, Chapter 7. Francis X. Diebold, Elements of Forecasting, 4th Edition (Mason, Ohio: Cengage Learning, 2006). Chapter 7. Characterizing Cycles

Diebold, Chapter 7. Francis X. Diebold, Elements of Forecasting, 4th Edition (Mason, Ohio: Cengage Learning, 2006). Chapter 7. Characterizing Cycles Diebold, Chaper 7 Francis X. Diebold, Elemens of Forecasing, 4h Ediion (Mason, Ohio: Cengage Learning, 006). Chaper 7. Characerizing Cycles Afer compleing his reading you should be able o: Define covariance

More information

Diffeomorphic Shape Momentum, Computational Anatomy, & Neuroinformatics at 1mm Scale

Diffeomorphic Shape Momentum, Computational Anatomy, & Neuroinformatics at 1mm Scale Diffeomorphic Shape Momenum, Compuaional Anaomy, & Neuroinformaics a 1mm Scale Michael I. Miller Torono, June 2011 Insiue for Compuaional Medicine Neuroinformaics a 1mm Scale Diffeomorphic Mapping Subcorical

More information

BEng (Hons) Telecommunications. Examinations for / Semester 2

BEng (Hons) Telecommunications. Examinations for / Semester 2 BEng (Hons) Telecommunicaions Cohor: BTEL/14/FT Examinaions for 2015-2016 / Semeser 2 MODULE: ELECTROMAGNETIC THEORY MODULE CODE: ASE2103 Duraion: 2 ½ Hours Insrucions o Candidaes: 1. Answer ALL 4 (FOUR)

More information

Diverse Nonnegative Matrix Factorization for Multi-view Data Representation

Diverse Nonnegative Matrix Factorization for Multi-view Data Representation Diverse Nonnegaive Marix Facorizaion for Muli-view Daa Represenaion Jing Wang, Feng Tian, Hongchuan Yu, Member, IEEE, Chang Hong Liu, Kun Zhan, and Xiao Wang Absrac Nonnegaive marix facorizaion (NMF),

More information

Active Dictionary Learning for Image Representation

Active Dictionary Learning for Image Representation Acive Dicionary Learning for Image Represenaion Tong Wu, Anand D. Sarwae, and Waheed U. Bajwa Deparmen of Elecrical and Compuer Engineering Rugers, The Sae Universiy of New Jersey, Piscaaway, NJ 08854

More information

Zürich. ETH Master Course: L Autonomous Mobile Robots Localization II

Zürich. ETH Master Course: L Autonomous Mobile Robots Localization II Roland Siegwar Margaria Chli Paul Furgale Marco Huer Marin Rufli Davide Scaramuzza ETH Maser Course: 151-0854-00L Auonomous Mobile Robos Localizaion II ACT and SEE For all do, (predicion updae / ACT),

More information

THE GENERALIZED PASCAL MATRIX VIA THE GENERALIZED FIBONACCI MATRIX AND THE GENERALIZED PELL MATRIX

THE GENERALIZED PASCAL MATRIX VIA THE GENERALIZED FIBONACCI MATRIX AND THE GENERALIZED PELL MATRIX J Korean Mah Soc 45 008, No, pp 479 49 THE GENERALIZED PASCAL MATRIX VIA THE GENERALIZED FIBONACCI MATRIX AND THE GENERALIZED PELL MATRIX Gwang-yeon Lee and Seong-Hoon Cho Reprined from he Journal of he

More information

6.003 Homework 1. Problems. Due at the beginning of recitation on Wednesday, February 10, 2010.

6.003 Homework 1. Problems. Due at the beginning of recitation on Wednesday, February 10, 2010. 6.003 Homework Due a he beginning of reciaion on Wednesday, February 0, 200. Problems. Independen and Dependen Variables Assume ha he heigh of a waer wave is given by g(x v) where x is disance, v is velociy,

More information

New Boosting Methods of Gaussian Processes for Regression

New Boosting Methods of Gaussian Processes for Regression New Boosing Mehods of Gaussian Processes for Regression Yangqiu Song Sae Key Laboraory of Inelligen echnology and Sysems Deparmen of Auomaion singhua Universiy Beijing 00084, P.R.China E-mail: songyq99@mails.singhua.edu.cn

More information

Finish reading Chapter 2 of Spivak, rereading earlier sections as necessary. handout and fill in some missing details!

Finish reading Chapter 2 of Spivak, rereading earlier sections as necessary. handout and fill in some missing details! MAT 257, Handou 6: Ocober 7-2, 20. I. Assignmen. Finish reading Chaper 2 of Spiva, rereading earlier secions as necessary. handou and fill in some missing deails! II. Higher derivaives. Also, read his

More information

Application of a Stochastic-Fuzzy Approach to Modeling Optimal Discrete Time Dynamical Systems by Using Large Scale Data Processing

Application of a Stochastic-Fuzzy Approach to Modeling Optimal Discrete Time Dynamical Systems by Using Large Scale Data Processing Applicaion of a Sochasic-Fuzzy Approach o Modeling Opimal Discree Time Dynamical Sysems by Using Large Scale Daa Processing AA WALASZE-BABISZEWSA Deparmen of Compuer Engineering Opole Universiy of Technology

More information

Section 4.4 Logarithmic Properties

Section 4.4 Logarithmic Properties Secion. Logarihmic Properies 59 Secion. Logarihmic Properies In he previous secion, we derived wo imporan properies of arihms, which allowed us o solve some asic eponenial and arihmic equaions. Properies

More information

Robert Kollmann. 6 September 2017

Robert Kollmann. 6 September 2017 Appendix: Supplemenary maerial for Tracable Likelihood-Based Esimaion of Non- Linear DSGE Models Economics Leers (available online 6 Sepember 207) hp://dx.doi.org/0.06/j.econle.207.08.027 Rober Kollmann

More information

RANDOM LAGRANGE MULTIPLIERS AND TRANSVERSALITY

RANDOM LAGRANGE MULTIPLIERS AND TRANSVERSALITY ECO 504 Spring 2006 Chris Sims RANDOM LAGRANGE MULTIPLIERS AND TRANSVERSALITY 1. INTRODUCTION Lagrange muliplier mehods are sandard fare in elemenary calculus courses, and hey play a cenral role in economic

More information

Notes on Kalman Filtering

Notes on Kalman Filtering Noes on Kalman Filering Brian Borchers and Rick Aser November 7, Inroducion Daa Assimilaion is he problem of merging model predicions wih acual measuremens of a sysem o produce an opimal esimae of he curren

More information

CSE/NB 528 Lecture 14: Reinforcement Learning (Chapter 9)

CSE/NB 528 Lecture 14: Reinforcement Learning (Chapter 9) CSE/NB 528 Lecure 14: Reinforcemen Learning Chaper 9 Image from hp://clasdean.la.asu.edu/news/images/ubep2001/neuron3.jpg Lecure figures are from Dayan & Abbo s book hp://people.brandeis.edu/~abbo/book/index.hml

More information

U( θ, θ), U(θ 1/2, θ + 1/2) and Cauchy (θ) are not exponential families. (The proofs are not easy and require measure theory. See the references.

U( θ, θ), U(θ 1/2, θ + 1/2) and Cauchy (θ) are not exponential families. (The proofs are not easy and require measure theory. See the references. Lecure 5 Exponenial Families Exponenial families, also called Koopman-Darmois families, include a quie number of well known disribuions. Many nice properies enjoyed by exponenial families allow us o provide

More information

Elements of Stochastic Processes Lecture II Hamid R. Rabiee

Elements of Stochastic Processes Lecture II Hamid R. Rabiee Sochasic Processes Elemens of Sochasic Processes Lecure II Hamid R. Rabiee Overview Reading Assignmen Chaper 9 of exbook Furher Resources MIT Open Course Ware S. Karlin and H. M. Taylor, A Firs Course

More information

A problem related to Bárány Grünbaum conjecture

A problem related to Bárány Grünbaum conjecture Filoma 27:1 (2013), 109 113 DOI 10.2298/FIL1301109B Published by Faculy of Sciences and Mahemaics, Universiy of Niš, Serbia Available a: hp://www.pmf.ni.ac.rs/filoma A problem relaed o Bárány Grünbaum

More information

1 Review of Zero-Sum Games

1 Review of Zero-Sum Games COS 5: heoreical Machine Learning Lecurer: Rob Schapire Lecure #23 Scribe: Eugene Brevdo April 30, 2008 Review of Zero-Sum Games Las ime we inroduced a mahemaical model for wo player zero-sum games. Any

More information

The Simple Linear Regression Model: Reporting the Results and Choosing the Functional Form

The Simple Linear Regression Model: Reporting the Results and Choosing the Functional Form Chaper 6 The Simple Linear Regression Model: Reporing he Resuls and Choosing he Funcional Form To complee he analysis of he simple linear regression model, in his chaper we will consider how o measure

More information

SUFFICIENT CONDITIONS FOR EXISTENCE SOLUTION OF LINEAR TWO-POINT BOUNDARY PROBLEM IN MINIMIZATION OF QUADRATIC FUNCTIONAL

SUFFICIENT CONDITIONS FOR EXISTENCE SOLUTION OF LINEAR TWO-POINT BOUNDARY PROBLEM IN MINIMIZATION OF QUADRATIC FUNCTIONAL HE PUBLISHING HOUSE PROCEEDINGS OF HE ROMANIAN ACADEMY, Series A, OF HE ROMANIAN ACADEMY Volume, Number 4/200, pp 287 293 SUFFICIEN CONDIIONS FOR EXISENCE SOLUION OF LINEAR WO-POIN BOUNDARY PROBLEM IN

More information

Chapter 15. Time Series: Descriptive Analyses, Models, and Forecasting

Chapter 15. Time Series: Descriptive Analyses, Models, and Forecasting Chaper 15 Time Series: Descripive Analyses, Models, and Forecasing Descripive Analysis: Index Numbers Index Number a number ha measures he change in a variable over ime relaive o he value of he variable

More information

Math 10B: Mock Mid II. April 13, 2016

Math 10B: Mock Mid II. April 13, 2016 Name: Soluions Mah 10B: Mock Mid II April 13, 016 1. ( poins) Sae, wih jusificaion, wheher he following saemens are rue or false. (a) If a 3 3 marix A saisfies A 3 A = 0, hen i canno be inverible. True.

More information

Math From Scratch Lesson 34: Isolating Variables

Math From Scratch Lesson 34: Isolating Variables Mah From Scrach Lesson 34: Isolaing Variables W. Blaine Dowler July 25, 2013 Conens 1 Order of Operaions 1 1.1 Muliplicaion and Addiion..................... 1 1.2 Division and Subracion.......................

More information

DIRICHLET-PROCESS-MIXTURE-BASED BAYESIAN NONPARAMETRIC METHOD FOR MARKOV SWITCHING PROCESS ESTIMATION

DIRICHLET-PROCESS-MIXTURE-BASED BAYESIAN NONPARAMETRIC METHOD FOR MARKOV SWITCHING PROCESS ESTIMATION 3rd European Signal Processing Conference (EUSIPCO) DIRICHLET-PROCESS-MIXTURE-BASED BAYESIAN NONPARAMETRIC METHOD FOR MARKOV SWITCHING PROCESS ESTIMATION Clémen Magnan [],[3] Audrey Giremus [3] Eric Grivel

More information

Advanced FDTD Algorithms

Advanced FDTD Algorithms EE 5303 Elecromagneic Analsis Using Finie Difference Time Domain Lecure #5 Advanced FDTD Algorihms Lecure 5 These noes ma conain coprighed maerial obained under fair use rules. Disribuion of hese maerials

More information

Laplace Transform and its Relation to Fourier Transform

Laplace Transform and its Relation to Fourier Transform Chaper 6 Laplace Transform and is Relaion o Fourier Transform (A Brief Summary) Gis of he Maer 2 Domains of Represenaion Represenaion of signals and sysems Time Domain Coninuous Discree Time Time () [n]

More information

An EM based training algorithm for recurrent neural networks

An EM based training algorithm for recurrent neural networks An EM based raining algorihm for recurren neural neworks Jan Unkelbach, Sun Yi, and Jürgen Schmidhuber IDSIA,Galleria 2, 6928 Manno, Swizerland {jan.unkelbach,yi,juergen}@idsia.ch hp://www.idsia.ch Absrac.

More information

5.1 - Logarithms and Their Properties

5.1 - Logarithms and Their Properties Chaper 5 Logarihmic Funcions 5.1 - Logarihms and Their Properies Suppose ha a populaion grows according o he formula P 10, where P is he colony size a ime, in hours. When will he populaion be 2500? We

More information

QUANTITATIVE DECAY FOR NONLINEAR WAVE EQUATIONS

QUANTITATIVE DECAY FOR NONLINEAR WAVE EQUATIONS QUANTITATIVE DECAY FOR NONLINEAR WAVE EQUATIONS SPUR FINAL PAPER, SUMMER 08 CALVIN HSU MENTOR: RUOXUAN YANG PROJECT SUGGESTED BY: ANDREW LAWRIE Augus, 08 Absrac. In his paper, we discuss he decay rae for

More information

ES.1803 Topic 22 Notes Jeremy Orloff

ES.1803 Topic 22 Notes Jeremy Orloff ES.83 Topic Noes Jeremy Orloff Fourier series inroducion: coninued. Goals. Be able o compue he Fourier coefficiens of even or odd periodic funcion using he simplified formulas.. Be able o wrie and graph

More information

8. Basic RL and RC Circuits

8. Basic RL and RC Circuits 8. Basic L and C Circuis This chaper deals wih he soluions of he responses of L and C circuis The analysis of C and L circuis leads o a linear differenial equaion This chaper covers he following opics

More information

Lecture Notes 2. The Hilbert Space Approach to Time Series

Lecture Notes 2. The Hilbert Space Approach to Time Series Time Series Seven N. Durlauf Universiy of Wisconsin. Basic ideas Lecure Noes. The Hilber Space Approach o Time Series The Hilber space framework provides a very powerful language for discussing he relaionship

More information

Finite Element Analysis of Structures

Finite Element Analysis of Structures KAIT OE5 Finie Elemen Analysis of rucures Mid-erm Exam, Fall 9 (p) m. As shown in Fig., we model a russ srucure of uniform area (lengh, Area Am ) subjeced o a uniform body force ( f B e x N / m ) using

More information

Article from. Predictive Analytics and Futurism. July 2016 Issue 13

Article from. Predictive Analytics and Futurism. July 2016 Issue 13 Aricle from Predicive Analyics and Fuurism July 6 Issue An Inroducion o Incremenal Learning By Qiang Wu and Dave Snell Machine learning provides useful ools for predicive analyics The ypical machine learning

More information

An Efficient Image Similarity Measure based on Approximations of KL-Divergence Between Two Gaussian Mixtures

An Efficient Image Similarity Measure based on Approximations of KL-Divergence Between Two Gaussian Mixtures An Efficien Image Similariy Measure based on Approximaions of KL-Divergence Beween Two Gaussian Mixures Jacob Goldberger Shiri Gordon Hayi Greenspan Cue Sysems Tel Aviv Israel The Engineering Deparmen

More information

3.1.3 INTRODUCTION TO DYNAMIC OPTIMIZATION: DISCRETE TIME PROBLEMS. A. The Hamiltonian and First-Order Conditions in a Finite Time Horizon

3.1.3 INTRODUCTION TO DYNAMIC OPTIMIZATION: DISCRETE TIME PROBLEMS. A. The Hamiltonian and First-Order Conditions in a Finite Time Horizon 3..3 INRODUCION O DYNAMIC OPIMIZAION: DISCREE IME PROBLEMS A. he Hamilonian and Firs-Order Condiions in a Finie ime Horizon Define a new funcion, he Hamilonian funcion, H. H he change in he oal value of

More information

AP Chemistry--Chapter 12: Chemical Kinetics

AP Chemistry--Chapter 12: Chemical Kinetics AP Chemisry--Chaper 12: Chemical Kineics I. Reacion Raes A. The area of chemisry ha deals wih reacion raes, or how fas a reacion occurs, is called chemical kineics. B. The rae of reacion depends on he

More information

EE3723 : Digital Communications

EE3723 : Digital Communications EE373 : Digial Communicaions Week 6-7: Deecion Error Probabiliy Signal Space Orhogonal Signal Space MAJU-Digial Comm.-Week-6-7 Deecion Mached filer reduces he received signal o a single variable zt, afer

More information