Learning with Partially Observed Data
|
|
- Derrick Underwood
- 5 years ago
- Views:
Transcription
1 Readngs K&F Learnng wth artall Observed ata Lecture 2 Ma 4 2 CSE 55 Statstcal Methods Sprng 2 Instructor Su-In Lee nverst of Washngton Seattle Model Selecton So far we focused on sngle model ~ ven {[] [M]} fnd best scorng model arg ma ~ se t to predct net eample [ M + ] Implct assumpton Makng predctons based on the Baesan estmaton rule [ M + ] [ M + ] Best scorng model domnates the weghted sum ~ [ M + ] [ M + ] Vald wth man data nstances ver large M ros We get a sngle structure Allows for effcent use n our predcton tasks Cons Commttng to the ndependences of a partcular structure Other structures wth smlar score mght be probable gven 2
2 Model Selecton enst estmaton ckng one structure ma suffce f t dstrbuton [ M + ] s smlar for dfferent hgh-scorng structures. Structure dscover Several networks wth smlar scores one or several of them mght be close to the true structure but we cannot dstngush between them gven the data. rawng a concluson about the structure from one of the networks can be wrong Thus nstead of pckng one of the hgh-scorng structures we should focus on estmatng the confdence of the structural propertes we are nterested n. efne features f e.g. edge sub-structure d-sep propert Compute f f Requres summng over eponentall man structures We can reduce the computaton assumng a certan orderng 3 f Model Averagng ven an Order g Assumptons Known total order of varables Mamum n-degree for varables d Margnal lkelhood f g g f g 2 + L+ f L f gm g n L f d d ep{ B a ep{ FamScore } B a } d d ep{ FamScoreB } { < ep FamScore < d} B g n { < < d} + L+ f gm n Cost per faml On d Total cost On d+ 2 L + L n sng decomposablt assumpton on pror { } Snce gven orderng parent choces are ndependent 4 n 2
3 3 Model Averagng ven an Order osteror probablt of a general feature f f partcular choce of parents for f estence of a partcular edge between j { } { } < < } { ep ep d B B FamScore FamScore a All terms cancel out { } < < d B d FamScore f f f } { ep { } { } < < < < } { } { ep ep d B d and B j j FamScore FamScore a 5 Model Averagng We cannot assume that order s known Soluton Sample from posteror dstrbuton of If we manage to sample graphs.. K from Estmate feature probablt b Samplng can be done b MCMC Markov chan Monte Carlo Net week K d f K f 6
4 Notes on Learnng Local Structures Beond table Cs efne score wth local structures Eample n tree Cs score decomposes b leaves not b and a partcular value on ar ror ma need to be etended Eample n tree Cs penalt for tree structure per C depth of the tree Etend search operators to local structure Eample n tree Cs we need to search for tree structure Can be done b local encapsulated search or b defnng new global operatons 7 Structure Search Summar screte optmzaton problem In general N-ard Need to resort to heurstc search In practce search s relatvel fast ~ vars n ~ mn ecomposablt Suffcent statstcs In some cases we can reduce the search problem to an eas optmzaton problem Eample learnng trees a fed orderng 8 4
5 Let s turn to the man topc for toda LEARNIN WIT ARTIALL OBSERVE ATA 9 Tranng ata Tranng nstance Θ <Θ > ISCONNECT 5 A SNT VENTLN VENITBE RESS MINOVL FIO2 VENTALV ANALAIS VSAT ARTCO2 TR SAO2 INSFFANEST ECO2 N- N OVOLEMIA LVFAILRE CATECOL LVEVOLME STROEVOLME ISTOR ERRBLOWOTT R ERRCATER CV CW CO REK RSAT RB B ntl now we assumed that the tranng data s full observed Each nstance assgns values to all the varables n our doman 5
6 Incomplete ata In realt ths assumpton mght not be true. Tranng nstance <Θ > Lung cancer? 3? -? 2 9 8? ? 8 2? ? A ANALAIS TR Θ 2 3 SNT MINOVL SAO2 FIO2 VSAT INSFFANEST VENTLN VENTALV ARTCO2 ECO2 RESS 4 5 VENITBE ISCONNECT N- N?????????????????? 7 4? 7 OVOLEMIA LVFAILRE CATECOL LVEVOLME STROEVOLME ISTOR ERRBLOWOTT R ERRCATER CV CW CO REK RSAT RB Mssng values dden varables B Challenges Foundatonal s the learnng task well defned? Computatonal how can we learn wth mssng data? Treatng Mssng ata ow should we treat mssng data? Based on data mssng mechansm Case I A con s tossed on a table occasonall t drops and measurements are not taken random mssng Sample sequence T??T? Treat mssng data b gnorng t Case II A con s tossed but onl heads are reported delberate mssng values Sample sequence???? Treat mssng data b fllng t wth Tals We need to consder the data mssng mechansm 2 6
7 Modelng ata Mssng Mechansm Let s tr to model the data mssng mechansm {... n } are random varables O {O...O n } are observablt varables Alwas observed {... n } new random varables Val Val {?} s a determnstc functon of and O O o? O o 3 Modelng Mssng ata Mechansm Case I random mssng values Case II delberate mssng values ψ ψ O O ψ T ψ? ψ L ψ M MT ψ M + MT ψ M? MLE M ˆ M + M M + M T ψˆ M + M + M T T 4? 7
8 Modelng Mssng ata Mechansm Case I random mssng values Case II delberate mssng values ψ O MLE? M ˆ M + M M + M T ψˆ M + M + M? T T? ψ O T? ψ O ψ ψ O T O + ψ O T M ψ + ψ? M MT M MT L ψ ψ O ψ O T O O T ecouplng of Observaton Mechansm When can we gnore the mssng data mechansm and focus onl on the lkelhood? Mssng Completel at Random MCAR For ever Ind ;O a ver strong assumpton Suffcent but not necessar for the decomposton of the lkelhood Mssng at Random MAR s suffcent The probablt that the value of s mssng s ndependent of ts actual value gven other observed values In both cases the lkelhood decomposes When there are mssng values n tr to model such that MAR holds. 6 8
9 Incomplete ata In realt ths assumpton mght not be true Lung cancer? 3? -? 2 9 8? ? 8 2? ? Θ <Θ > ISCONNECT 5 A SNT VENTLN VENITBE RESS MINOVL FIO2 VENTALV ANALAIS VSAT ARTCO2 TR SAO2 INSFFANEST ECO2 N- N?????????????????? 7 4? 7 OVOLEMIA LVFAILRE CATECOL LVEVOLME STROEVOLME ISTOR ERRBLOWOTT R ERRCATER CV CW CO REK RSAT RB Mssng values dden varables B Challenges Foundatonal s the learnng task well defned? Computatonal how can we learn wth mssng data? 7 dden Latent Varables Attempt to learn a model wth hdden varables In ths case MCAR alwas holds varable s alwas mssng Wh should we care about unobserved varables? parameters 59 parameters 8 9
10 dden Latent Varables dden varables also appear n clusterng Naïve Baes model Class varable s hdden Observed attrbutes are ndependent gven the class N- N Cluster dden 2... Observed possble mssng values n ow do mssng data affect the lkelhood functon? 2
11 Lkelhood for Complete ata [3] [3] [2] [2] [] [] L Input ata Lkelhood Lkelhood decomposes b varables Lkelhood decomposes wthn Cs Lkelhood functon s log-concave unque global mamum that has a smple analtc closed form. 2 Lkelhood for Incomplete ata?? 2 L Input ata Lkelhood Lkelhood does not decompose b varables Lkelhood does not decompose wthn Cs Computng lkelhood per nstance requres nference! 22
12 Lkelhood wth Mssng ata Multmodal lkelhood functon wth ncomplete data Lkelhood functon s not log-concave local mama cannot be obtaned b a smple analtc closed form CSE 55 Statstcal Methods Sprng 2 23 MLE from Incomplete ata Take steps proportonal to the postve of the gradent. LΘ radent Ascent Follow gradent of lkelhood w.r.t. to parameters Add lne search and conjugate gradent methods to get fast convergence Θ 24 2
13 MLE from Incomplete ata Nonlnear optmzaton problem LΘ Θ Epectaton Mamzaton EM se current pont to construct alternatve functon whch s nce uarant mamum of new functon has better score than current pont 25 MLE from Incomplete ata Nonlnear optmzaton problem LΘ radent Ascent and EM Fnd local mama Requre multple restarts to fnd appro. to the global mamum Requre computatons n each teraton Θ 26 3
14 radent Ascent Theorem log Θ pa roof log Θ pa m m pa m pa log m] Θ pa m] Θ m] Θ m] Θ pa a T Θ Θ T T Θ T T Θ T T a observed data n the m-th nstance each tranng nstance ow do we compute? m] Θ pa 27 radent Ascent a m] Θ pa pa Θ < O> m] pa pa m] Θ Θ < O> m] < a >< pa > pa log Θ pa m m log m] Θ pa m] Θ m] Θ pa 28 4
15 radent Ascent log Θ pa m m m m] Θ m] Θ pa m] Θ m] Θ pa m] Θ pa pa pa a Requres computaton pa m]θ for all m Can be done wth clque-tree algorthm snce a are n the same clque 29 radent Ascent Summar ros Fleble can be etended to non table Cs Cons Need to project gradent onto space of legal parameters For reasonable convergence need to combne wth advanced methods conjugate gradent lne search 3 5
16 Epectaton Mamzaton EM Talored algorthm for optmzng lkelhood functons Intuton arameter estmaton s eas gven complete data Computng probablt of mssng data s eas nference gven parameters Strateg ck a startng pont for parameters Complete the data usng current parameters Estmate parameters relatve to data completon Iterate rocedure guaranteed to mprove at each teraton 3 Epectaton Mamzaton EM Intalze parameters to Iterate E-step and M-step In the t-th teraton we do Epectaton E-step Let m] be the observed data n the m-th tranng nstance. For each m and each faml a compute a m] t Compute the epected suffcent statstcs for each values u on a respectvel. t M t [ a u] a u m] Mamzaton M-step Treat the epected suffcent statstcs as observed and set the parameters to the MLE wth respect to the ESS M [ a u] t+ t a u M [ a u] m t 32 6
17 Epectaton Mamzaton EM Intal network pdated network Epected counts N + E-Step nference N M-Step reparameterze Tranng data?? Iterate 33 Epectaton Mamzaton EM Formal uarantees LΘ t+ LΘ t Each teraton mproves the lkelhood If Θ t+ Θ t then Θ t s a statonar pont of LΘ suall ths means a local mamum Man cost Computatons of epected counts n E-Step Requres nference for each nstance n tranng set Eactl the same as n gradent ascent! Readng materal on EM lease read Andrew Ng s lecture note 34 7
18 EM ractcal Consderatons Intal parameters ghl senstve to startng parameters Choose randoml Choose b guessng from another source Stoppng crtera Small change n data lkelhood Small change n parameters Avodng bad local mama Multple restarts Earl prunng of unpromsng startng ponts 35 Acknowledgement These lecture notes were generated based on the sldes from rof Eran Segal. CSE 55 Statstcal Methods Sprng
Expectation Maximization Mixture Models HMMs
-755 Machne Learnng for Sgnal Processng Mture Models HMMs Class 9. 2 Sep 200 Learnng Dstrbutons for Data Problem: Gven a collecton of eamples from some data, estmate ts dstrbuton Basc deas of Mamum Lelhood
More informationMachine learning: Density estimation
CS 70 Foundatons of AI Lecture 3 Machne learnng: ensty estmaton Mlos Hauskrecht mlos@cs.ptt.edu 539 Sennott Square ata: ensty estmaton {.. n} x a vector of attrbute values Objectve: estmate the model of
More informationCourse 395: Machine Learning - Lectures
Course 395: Machne Learnng - Lectures Lecture 1-2: Concept Learnng (M. Pantc Lecture 3-4: Decson Trees & CC Intro (M. Pantc Lecture 5-6: Artfcal Neural Networks (S.Zaferou Lecture 7-8: Instance ased Learnng
More informationCIS526: Machine Learning Lecture 3 (Sept 16, 2003) Linear Regression. Preparation help: Xiaoying Huang. x 1 θ 1 output... θ M x M
CIS56: achne Learnng Lecture 3 (Sept 6, 003) Preparaton help: Xaoyng Huang Lnear Regresson Lnear regresson can be represented by a functonal form: f(; θ) = θ 0 0 +θ + + θ = θ = 0 ote: 0 s a dummy attrbute
More informationEM and Structure Learning
EM and Structure Learnng Le Song Machne Learnng II: Advanced Topcs CSE 8803ML, Sprng 2012 Partally observed graphcal models Mxture Models N(μ 1, Σ 1 ) Z X N N(μ 2, Σ 2 ) 2 Gaussan mxture model Consder
More informationStructure Learning. Instructor: Su-In Lee University of Washington, Seattle. Score-based structure learning
Readngs: K&F 18.3, 18.4, 18.5, 18.6 Structure Learnng Lecture 11 ay 2, 2011 SE 515, Statstcal ethods, Sprng 2011 Instructor: Su-In Lee Unversty of Washngton, Seattle Last Tme Score-based structure learnng
More informationParametric fractional imputation for missing data analysis. Jae Kwang Kim Survey Working Group Seminar March 29, 2010
Parametrc fractonal mputaton for mssng data analyss Jae Kwang Km Survey Workng Group Semnar March 29, 2010 1 Outlne Introducton Proposed method Fractonal mputaton Approxmaton Varance estmaton Multple mputaton
More informationAn Experiment/Some Intuition (Fall 2006): Lecture 18 The EM Algorithm heads coin 1 tails coin 2 Overview Maximum Likelihood Estimation
An Experment/Some Intuton I have three cons n my pocket, 6.864 (Fall 2006): Lecture 18 The EM Algorthm Con 0 has probablty λ of heads; Con 1 has probablty p 1 of heads; Con 2 has probablty p 2 of heads
More informationMLE and Bayesian Estimation. Jie Tang Department of Computer Science & Technology Tsinghua University 2012
MLE and Bayesan Estmaton Je Tang Department of Computer Scence & Technology Tsnghua Unversty 01 1 Lnear Regresson? As the frst step, we need to decde how we re gong to represent the functon f. One example:
More informationGenerative classification models
CS 675 Intro to Machne Learnng Lecture Generatve classfcaton models Mlos Hauskrecht mlos@cs.ptt.edu 539 Sennott Square Data: D { d, d,.., dn} d, Classfcaton represents a dscrete class value Goal: learn
More informationMean Field / Variational Approximations
Mean Feld / Varatonal Appromatons resented by Jose Nuñez 0/24/05 Outlne Introducton Mean Feld Appromaton Structured Mean Feld Weghted Mean Feld Varatonal Methods Introducton roblem: We have dstrbuton but
More informationLogistic Regression. CAP 5610: Machine Learning Instructor: Guo-Jun QI
Logstc Regresson CAP 561: achne Learnng Instructor: Guo-Jun QI Bayes Classfer: A Generatve model odel the posteror dstrbuton P(Y X) Estmate class-condtonal dstrbuton P(X Y) for each Y Estmate pror dstrbuton
More informationCS 2750 Machine Learning. Lecture 5. Density estimation. CS 2750 Machine Learning. Announcements
CS 750 Machne Learnng Lecture 5 Densty estmaton Mlos Hauskrecht mlos@cs.ptt.edu 539 Sennott Square CS 750 Machne Learnng Announcements Homework Due on Wednesday before the class Reports: hand n before
More information1. Inference on Regression Parameters a. Finding Mean, s.d and covariance amongst estimates. 2. Confidence Intervals and Working Hotelling Bands
Content. Inference on Regresson Parameters a. Fndng Mean, s.d and covarance amongst estmates.. Confdence Intervals and Workng Hotellng Bands 3. Cochran s Theorem 4. General Lnear Testng 5. Measures of
More informationDiscriminative classifier: Logistic Regression. CS534-Machine Learning
Dscrmnatve classfer: Logstc Regresson CS534-Machne Learnng 2 Logstc Regresson Gven tranng set D stc regresson learns the condtonal dstrbuton We ll assume onl to classes and a parametrc form for here s
More informationHidden Markov Models
CM229S: Machne Learnng for Bonformatcs Lecture 12-05/05/2016 Hdden Markov Models Lecturer: Srram Sankararaman Scrbe: Akshay Dattatray Shnde Edted by: TBD 1 Introducton For a drected graph G we can wrte
More informationFinite Mixture Models and Expectation Maximization. Most slides are from: Dr. Mario Figueiredo, Dr. Anil Jain and Dr. Rong Jin
Fnte Mxture Models and Expectaton Maxmzaton Most sldes are from: Dr. Maro Fgueredo, Dr. Anl Jan and Dr. Rong Jn Recall: The Supervsed Learnng Problem Gven a set of n samples X {(x, y )},,,n Chapter 3 of
More informationHomework Assignment 3 Due in class, Thursday October 15
Homework Assgnment 3 Due n class, Thursday October 15 SDS 383C Statstcal Modelng I 1 Rdge regresson and Lasso 1. Get the Prostrate cancer data from http://statweb.stanford.edu/~tbs/elemstatlearn/ datasets/prostate.data.
More informationEnsemble Methods: Boosting
Ensemble Methods: Boostng Ncholas Ruozz Unversty of Texas at Dallas Based on the sldes of Vbhav Gogate and Rob Schapre Last Tme Varance reducton va baggng Generate new tranng data sets by samplng wth replacement
More informationSemi-Supervised Learning
Sem-Supervsed Learnng Consder the problem of Prepostonal Phrase Attachment. Buy car wth money ; buy car wth wheel There are several ways to generate features. Gven the lmted representaton, we can assume
More informationClassification learning II
Lecture 8 Classfcaton learnng II Mlos Hauskrecht mlos@cs.ptt.edu 539 Sennott Square Logstc regresson model Defnes a lnear decson boundar Dscrmnant functons: g g g g here g z / e z f, g g - s a logstc functon
More informationINF 5860 Machine learning for image classification. Lecture 3 : Image classification and regression part II Anne Solberg January 31, 2018
INF 5860 Machne learnng for mage classfcaton Lecture 3 : Image classfcaton and regresson part II Anne Solberg January 3, 08 Today s topcs Multclass logstc regresson and softma Regularzaton Image classfcaton
More informationLecture Nov
Lecture 18 Nov 07 2008 Revew Clusterng Groupng smlar obects nto clusters Herarchcal clusterng Agglomeratve approach (HAC: teratvely merge smlar clusters Dfferent lnkage algorthms for computng dstances
More informationMachine Learning. Learning Bayesian networks. The Learning Problem
Machne Learnng Vasant onavar rtfcal Intellgence Research Laboratory epartment of Computer Scence Computatonal Intellgence, Learnng, & scovery rogram Iowa State Unversty honavar@csastateedu wwwcsastateedu/~honavar
More informationCHAPTER 3: BAYESIAN DECISION THEORY
HATER 3: BAYESIAN DEISION THEORY Decson mang under uncertanty 3 Data comes from a process that s completely not nown The lac of nowledge can be compensated by modelng t as a random process May be the underlyng
More informationxp(x µ) = 0 p(x = 0 µ) + 1 p(x = 1 µ) = µ
CSE 455/555 Sprng 2013 Homework 7: Parametrc Technques Jason J. Corso Computer Scence and Engneerng SUY at Buffalo jcorso@buffalo.edu Solutons by Yngbo Zhou Ths assgnment does not need to be submtted and
More informationDiscriminative classifier: Logistic Regression. CS534-Machine Learning
Dscrmnatve classfer: Logstc Regresson CS534-Machne Learnng robablstc Classfer Gven an nstance, hat does a probablstc classfer do dfferentl compared to, sa, perceptron? It does not drectl predct Instead,
More information9 : Learning Partially Observed GM : EM Algorithm
10-708: Probablstc Graphcal Models 10-708, Sprng 2014 9 : Learnng Partally Observed GM : EM Algorthm Lecturer: Erc P. Xng Scrbes: Rohan Ramanath, Rahul Goutam 1 Generalzed Iteratve Scalng In ths secton,
More informationFeature Selection: Part 1
CSE 546: Machne Learnng Lecture 5 Feature Selecton: Part 1 Instructor: Sham Kakade 1 Regresson n the hgh dmensonal settng How do we learn when the number of features d s greater than the sample sze n?
More informationMaximum Likelihood Estimation of Binary Dependent Variables Models: Probit and Logit. 1. General Formulation of Binary Dependent Variables Models
ECO 452 -- OE 4: Probt and Logt Models ECO 452 -- OE 4 Mamum Lkelhood Estmaton of Bnary Dependent Varables Models: Probt and Logt hs note demonstrates how to formulate bnary dependent varables models for
More informationMixture o f of Gaussian Gaussian clustering Nov
Mture of Gaussan clusterng Nov 11 2009 Soft vs hard lusterng Kmeans performs Hard clusterng: Data pont s determnstcally assgned to one and only one cluster But n realty clusters may overlap Soft-clusterng:
More informationEvaluation for sets of classes
Evaluaton for Tet Categorzaton Classfcaton accuracy: usual n ML, the proporton of correct decsons, Not approprate f the populaton rate of the class s low Precson, Recall and F 1 Better measures 21 Evaluaton
More informationBayesian Learning. Smart Home Health Analytics Spring Nirmalya Roy Department of Information Systems University of Maryland Baltimore County
Smart Home Health Analytcs Sprng 2018 Bayesan Learnng Nrmalya Roy Department of Informaton Systems Unversty of Maryland Baltmore ounty www.umbc.edu Bayesan Learnng ombnes pror knowledge wth evdence to
More informatione i is a random error
Chapter - The Smple Lnear Regresson Model The lnear regresson equaton s: where + β + β e for,..., and are observable varables e s a random error How can an estmaton rule be constructed for the unknown
More informationKernel Methods and SVMs Extension
Kernel Methods and SVMs Extenson The purpose of ths document s to revew materal covered n Machne Learnng 1 Supervsed Learnng regardng support vector machnes (SVMs). Ths document also provdes a general
More informationRockefeller College University at Albany
Rockefeller College Unverst at Alban PAD 705 Handout: Maxmum Lkelhood Estmaton Orgnal b Davd A. Wse John F. Kenned School of Government, Harvard Unverst Modfcatons b R. Karl Rethemeer Up to ths pont n
More informationMarkov Chain Monte Carlo (MCMC), Gibbs Sampling, Metropolis Algorithms, and Simulated Annealing Bioinformatics Course Supplement
Markov Chan Monte Carlo MCMC, Gbbs Samplng, Metropols Algorthms, and Smulated Annealng 2001 Bonformatcs Course Supplement SNU Bontellgence Lab http://bsnuackr/ Outlne! Markov Chan Monte Carlo MCMC! Metropols-Hastngs
More informationWhich Separator? Spring 1
Whch Separator? 6.034 - Sprng 1 Whch Separator? Mamze the margn to closest ponts 6.034 - Sprng Whch Separator? Mamze the margn to closest ponts 6.034 - Sprng 3 Margn of a pont " # y (w $ + b) proportonal
More informationLecture 20: Hypothesis testing
Lecture : Hpothess testng Much of statstcs nvolves hpothess testng compare a new nterestng hpothess, H (the Alternatve hpothess to the borng, old, well-known case, H (the Null Hpothess or, decde whether
More informationChapter 2 - The Simple Linear Regression Model S =0. e i is a random error. S β2 β. This is a minimization problem. Solution is a calculus exercise.
Chapter - The Smple Lnear Regresson Model The lnear regresson equaton s: where y + = β + β e for =,..., y and are observable varables e s a random error How can an estmaton rule be constructed for the
More informationClassification as a Regression Problem
Target varable y C C, C,, ; Classfcaton as a Regresson Problem { }, 3 L C K To treat classfcaton as a regresson problem we should transform the target y nto numercal values; The choce of numercal class
More informationSpeech and Language Processing
Speech and Language rocessng Lecture 3 ayesan network and ayesan nference Informaton and ommuncatons Engneerng ourse Takahro Shnozak 08//5 Lecture lan (Shnozak s part) I gves the frst 6 lectures about
More informationEconomics 130. Lecture 4 Simple Linear Regression Continued
Economcs 130 Lecture 4 Contnued Readngs for Week 4 Text, Chapter and 3. We contnue wth addressng our second ssue + add n how we evaluate these relatonshps: Where do we get data to do ths analyss? How do
More informationMarkov Chain Monte Carlo Lecture 6
where (x 1,..., x N ) X N, N s called the populaton sze, f(x) f (x) for at least one {1, 2,..., N}, and those dfferent from f(x) are called the tral dstrbutons n terms of mportance samplng. Dfferent ways
More informationGeneralized Linear Methods
Generalzed Lnear Methods 1 Introducton In the Ensemble Methods the general dea s that usng a combnaton of several weak learner one could make a better learner. More formally, assume that we have a set
More informationMachine Learning for Signal Processing Linear Gaussian Models
Machne Learnng for Sgnal Processng Lnear Gaussan Models Class 7. 30 Oct 204 Instructor: Bhksha Raj 755/8797 Recap: MAP stmators MAP (Mamum A Posteror: Fnd a best guess for (statstcall, gven knon = argma
More informationOn an Extension of Stochastic Approximation EM Algorithm for Incomplete Data Problems. Vahid Tadayon 1
On an Extenson of Stochastc Approxmaton EM Algorthm for Incomplete Data Problems Vahd Tadayon Abstract: The Stochastc Approxmaton EM (SAEM algorthm, a varant stochastc approxmaton of EM, s a versatle tool
More informationLecture Notes on Linear Regression
Lecture Notes on Lnear Regresson Feng L fl@sdueducn Shandong Unversty, Chna Lnear Regresson Problem In regresson problem, we am at predct a contnuous target value gven an nput feature vector We assume
More informationDr. Shalabh Department of Mathematics and Statistics Indian Institute of Technology Kanpur
Analyss of Varance and Desgn of Experment-I MODULE VIII LECTURE - 34 ANALYSIS OF VARIANCE IN RANDOM-EFFECTS MODEL AND MIXED-EFFECTS EFFECTS MODEL Dr Shalabh Department of Mathematcs and Statstcs Indan
More informationWeek 5: Neural Networks
Week 5: Neural Networks Instructor: Sergey Levne Neural Networks Summary In the prevous lecture, we saw how we can construct neural networks by extendng logstc regresson. Neural networks consst of multple
More informationLearning undirected Models. Instructor: Su-In Lee University of Washington, Seattle. Mean Field Approximation
Readngs: K&F 0.3, 0.4, 0.6, 0.7 Learnng undrected Models Lecture 8 June, 0 CSE 55, Statstcal Methods, Sprng 0 Instructor: Su-In Lee Unversty of Washngton, Seattle Mean Feld Approxmaton Is the energy functonal
More informationMIMA Group. Chapter 2 Bayesian Decision Theory. School of Computer Science and Technology, Shandong University. Xin-Shun SDU
Group M D L M Chapter Bayesan Decson heory Xn-Shun Xu @ SDU School of Computer Scence and echnology, Shandong Unversty Bayesan Decson heory Bayesan decson theory s a statstcal approach to data mnng/pattern
More information18.1 Introduction and Recap
CS787: Advanced Algorthms Scrbe: Pryananda Shenoy and Shjn Kong Lecturer: Shuch Chawla Topc: Streamng Algorthmscontnued) Date: 0/26/2007 We contnue talng about streamng algorthms n ths lecture, ncludng
More informationWhy Monte Carlo Integration? Introduction to Monte Carlo Method. Continuous Probability. Continuous Probability
Introducton to Monte Carlo Method Kad Bouatouch IRISA Emal: kad@rsa.fr Wh Monte Carlo Integraton? To generate realstc lookng mages, we need to solve ntegrals of or hgher dmenson Pel flterng and lens smulaton
More informationHidden Markov Models
Hdden Markov Models Namrata Vaswan, Iowa State Unversty Aprl 24, 204 Hdden Markov Model Defntons and Examples Defntons:. A hdden Markov model (HMM) refers to a set of hdden states X 0, X,..., X t,...,
More informationSpace of ML Problems. CSE 473: Artificial Intelligence. Parameter Estimation and Bayesian Networks. Learning Topics
/7/7 CSE 73: Artfcal Intellgence Bayesan - Learnng Deter Fox Sldes adapted from Dan Weld, Jack Breese, Dan Klen, Daphne Koller, Stuart Russell, Andrew Moore & Luke Zettlemoyer What s Beng Learned? Space
More informationCell Biology. Lecture 1: 10-Oct-12. Marco Grzegorczyk. (Gen-)Regulatory Network. Microarray Chips. (Gen-)Regulatory Network. (Gen-)Regulatory Network
5.0.202 Genetsche Netzwerke Wntersemester 202/203 ell ology Lecture : 0-Oct-2 Marco Grzegorczyk Gen-Regulatory Network Mcroarray hps G G 2 G 3 2 3 metabolte metabolte Gen-Regulatory Network Gen-Regulatory
More informationMACHINE APPLIED MACHINE LEARNING LEARNING. Gaussian Mixture Regression
11 MACHINE APPLIED MACHINE LEARNING LEARNING MACHINE LEARNING Gaussan Mture Regresson 22 MACHINE APPLIED MACHINE LEARNING LEARNING Bref summary of last week s lecture 33 MACHINE APPLIED MACHINE LEARNING
More informationSupport Vector Machines
CS 2750: Machne Learnng Support Vector Machnes Prof. Adrana Kovashka Unversty of Pttsburgh February 17, 2016 Announcement Homework 2 deadlne s now 2/29 We ll have covered everythng you need today or at
More informationSingular Value Decomposition: Theory and Applications
Sngular Value Decomposton: Theory and Applcatons Danel Khashab Sprng 2015 Last Update: March 2, 2015 1 Introducton A = UDV where columns of U and V are orthonormal and matrx D s dagonal wth postve real
More informationLecture 4. Instructor: Haipeng Luo
Lecture 4 Instructor: Hapeng Luo In the followng lectures, we focus on the expert problem and study more adaptve algorthms. Although Hedge s proven to be worst-case optmal, one may wonder how well t would
More informationMachine Learning for Signal Processing Linear Gaussian Models
Machne Learnng for Sgnal rocessng Lnear Gaussan Models lass 2. 2 Nov 203 Instructor: Bhsha Raj 2 Nov 203 755/8797 HW3 s up. Admnstrva rojects please send us an update 2 Nov 203 755/8797 2 Recap: MA stmators
More informationStatistics for Economics & Business
Statstcs for Economcs & Busness Smple Lnear Regresson Learnng Objectves In ths chapter, you learn: How to use regresson analyss to predct the value of a dependent varable based on an ndependent varable
More informationResource Allocation and Decision Analysis (ECON 8010) Spring 2014 Foundations of Regression Analysis
Resource Allocaton and Decson Analss (ECON 800) Sprng 04 Foundatons of Regresson Analss Readng: Regresson Analss (ECON 800 Coursepak, Page 3) Defntons and Concepts: Regresson Analss statstcal technques
More informationMATH 829: Introduction to Data Mining and Analysis The EM algorithm (part 2)
1/16 MATH 829: Introducton to Data Mnng and Analyss The EM algorthm (part 2) Domnque Gullot Departments of Mathematcal Scences Unversty of Delaware Aprl 20, 2016 Recall 2/16 We are gven ndependent observatons
More informationSDMML HT MSc Problem Sheet 4
SDMML HT 06 - MSc Problem Sheet 4. The recever operatng characterstc ROC curve plots the senstvty aganst the specfcty of a bnary classfer as the threshold for dscrmnaton s vared. Let the data space be
More informationCSC321 Tutorial 9: Review of Boltzmann machines and simulated annealing
CSC321 Tutoral 9: Revew of Boltzmann machnes and smulated annealng (Sldes based on Lecture 16-18 and selected readngs) Yue L Emal: yuel@cs.toronto.edu Wed 11-12 March 19 Fr 10-11 March 21 Outlne Boltzmann
More informationINTRODUCTION TO MACHINE LEARNING 3RD EDITION
ETHEM ALPAYDIN The MIT Press, 2014 Lecture Sldes for INTRODUCTION TO MACHINE LEARNING 3RD EDITION alpaydn@boun.edu.tr http://www.cmpe.boun.edu.tr/~ethem/2ml3e CHAPTER 3: BAYESIAN DECISION THEORY Probablty
More informationStat 543 Exam 2 Spring 2016
Stat 543 Exam 2 Sprng 206 I have nether gven nor receved unauthorzed assstance on ths exam. Name Sgned Date Name Prnted Ths Exam conssts of questons. Do at least 0 of the parts of the man exam. I wll score
More informationDepartment of Computer Science Artificial Intelligence Research Laboratory. Iowa State University MACHINE LEARNING
MACHINE LEANING Vasant Honavar Bonformatcs and Computatonal Bology rogram Center for Computatonal Intellgence, Learnng, & Dscovery Iowa State Unversty honavar@cs.astate.edu www.cs.astate.edu/~honavar/
More informationCorrelation and Regression. Correlation 9.1. Correlation. Chapter 9
Chapter 9 Correlaton and Regresson 9. Correlaton Correlaton A correlaton s a relatonshp between two varables. The data can be represented b the ordered pars (, ) where s the ndependent (or eplanator) varable,
More informationStatistical analysis using matlab. HY 439 Presented by: George Fortetsanakis
Statstcal analyss usng matlab HY 439 Presented by: George Fortetsanaks Roadmap Probablty dstrbutons Statstcal estmaton Fttng data to probablty dstrbutons Contnuous dstrbutons Contnuous random varable X
More informationChapter 14 Simple Linear Regression
Chapter 4 Smple Lnear Regresson Chapter 4 - Smple Lnear Regresson Manageral decsons often are based on the relatonshp between two or more varables. Regresson analss can be used to develop an equaton showng
More information10-701/ Machine Learning, Fall 2005 Homework 3
10-701/15-781 Machne Learnng, Fall 2005 Homework 3 Out: 10/20/05 Due: begnnng of the class 11/01/05 Instructons Contact questons-10701@autonlaborg for queston Problem 1 Regresson and Cross-valdaton [40
More informationPGM Learning Tasks and Metrics
Probablstc Graphcal odels Learnng Overvew PG Learnng Tasks and etrcs Learnng doan epert True dstrbuton P* aybe correspondng to a PG * dataset of nstances D{d],...d]} sapled fro P* elctaton Network Learnng
More informationDepartment of Quantitative Methods & Information Systems. Time Series and Their Components QMIS 320. Chapter 6
Department of Quanttatve Methods & Informaton Systems Tme Seres and Ther Components QMIS 30 Chapter 6 Fall 00 Dr. Mohammad Zanal These sldes were modfed from ther orgnal source for educatonal purpose only.
More informationChapter 9: Statistical Inference and the Relationship between Two Variables
Chapter 9: Statstcal Inference and the Relatonshp between Two Varables Key Words The Regresson Model The Sample Regresson Equaton The Pearson Correlaton Coeffcent Learnng Outcomes After studyng ths chapter,
More informationArtificial Intelligence Bayesian Networks
Artfcal Intellgence Bayesan Networks Adapted from sldes by Tm Fnn and Mare desjardns. Some materal borrowed from Lse Getoor. 1 Outlne Bayesan networks Network structure Condtonal probablty tables Condtonal
More informationEcon107 Applied Econometrics Topic 3: Classical Model (Studenmund, Chapter 4)
I. Classcal Assumptons Econ7 Appled Econometrcs Topc 3: Classcal Model (Studenmund, Chapter 4) We have defned OLS and studed some algebrac propertes of OLS. In ths topc we wll study statstcal propertes
More information6 Supplementary Materials
6 Supplementar Materals 61 Proof of Theorem 31 Proof Let m Xt z 1:T : l m Xt X,z 1:t Wethenhave mxt z1:t ˆm HX Xt z 1:T mxt z1:t m HX Xt z 1:T + mxt z 1:T HX We consder each of the two terms n equaton
More informationLogistic Regression Maximum Likelihood Estimation
Harvard-MIT Dvson of Health Scences and Technology HST.951J: Medcal Decson Support, Fall 2005 Instructors: Professor Lucla Ohno-Machado and Professor Staal Vnterbo 6.873/HST.951 Medcal Decson Support Fall
More informationEstimation: Part 2. Chapter GREG estimation
Chapter 9 Estmaton: Part 2 9. GREG estmaton In Chapter 8, we have seen that the regresson estmator s an effcent estmator when there s a lnear relatonshp between y and x. In ths chapter, we generalzed the
More informationStat 543 Exam 2 Spring 2016
Stat 543 Exam 2 Sprng 2016 I have nether gven nor receved unauthorzed assstance on ths exam. Name Sgned Date Name Prnted Ths Exam conssts of 11 questons. Do at least 10 of the 11 parts of the man exam.
More informationLimited Dependent Variables
Lmted Dependent Varables. What f the left-hand sde varable s not a contnuous thng spread from mnus nfnty to plus nfnty? That s, gven a model = f (, β, ε, where a. s bounded below at zero, such as wages
More informationParametric fractional imputation for missing data analysis
Secton on Survey Research Methods JSM 2008 Parametrc fractonal mputaton for mssng data analyss Jae Kwang Km Wayne Fuller Abstract Under a parametrc model for mssng data, the EM algorthm s a popular tool
More informationThe EM Algorithm (Dempster, Laird, Rubin 1977) The missing data or incomplete data setting: ODL(φ;Y ) = [Y;φ] = [Y X,φ][X φ] = X
The EM Algorthm (Dempster, Lard, Rubn 1977 The mssng data or ncomplete data settng: An Observed Data Lkelhood (ODL that s a mxture or ntegral of Complete Data Lkelhoods (CDL. (1a ODL(;Y = [Y;] = [Y,][
More informationProperties of Least Squares
Week 3 3.1 Smple Lnear Regresson Model 3. Propertes of Least Squares Estmators Y Y β 1 + β X + u weekly famly expendtures X weekly famly ncome For a gven level of x, the expected level of food expendtures
More informationProbabilistic Graphical Models
School of Computer Scence robablstc Graphcal Models Appromate Inference: Markov Chan Monte Carlo 05 07 Erc Xng Lecture 7 March 9 04 X X 075 05 05 03 X 3 Erc Xng @ CMU 005-04 Recap of Monte Carlo Monte
More informationExplaining the Stein Paradox
Explanng the Sten Paradox Kwong Hu Yung 1999/06/10 Abstract Ths report offers several ratonale for the Sten paradox. Sectons 1 and defnes the multvarate normal mean estmaton problem and ntroduces Sten
More informationMaximum Likelihood Estimation of Binary Dependent Variables Models: Probit and Logit. 1. General Formulation of Binary Dependent Variables Models
ECO 452 -- OE 4: Probt and Logt Models ECO 452 -- OE 4 Maxmum Lkelhood Estmaton of Bnary Dependent Varables Models: Probt and Logt hs note demonstrates how to formulate bnary dependent varables models
More informationThe Gaussian classifier. Nuno Vasconcelos ECE Department, UCSD
he Gaussan classfer Nuno Vasconcelos ECE Department, UCSD Bayesan decson theory recall that we have state of the world X observatons g decson functon L[g,y] loss of predctng y wth g Bayes decson rule s
More informationCS 3710: Visual Recognition Classification and Detection. Adriana Kovashka Department of Computer Science January 13, 2015
CS 3710: Vsual Recognton Classfcaton and Detecton Adrana Kovashka Department of Computer Scence January 13, 2015 Plan for Today Vsual recognton bascs part 2: Classfcaton and detecton Adrana s research
More informationLecture 4 Hypothesis Testing
Lecture 4 Hypothess Testng We may wsh to test pror hypotheses about the coeffcents we estmate. We can use the estmates to test whether the data rejects our hypothess. An example mght be that we wsh to
More informationImage classification. Given the bag-of-features representations of images from different classes, how do we learn a model for distinguishing i them?
Image classfcaton Gven te bag-of-features representatons of mages from dfferent classes ow do we learn a model for dstngusng tem? Classfers Learn a decson rule assgnng bag-offeatures representatons of
More informationLinear Approximation with Regularization and Moving Least Squares
Lnear Approxmaton wth Regularzaton and Movng Least Squares Igor Grešovn May 007 Revson 4.6 (Revson : March 004). 5 4 3 0.5 3 3.5 4 Contents: Lnear Fttng...4. Weghted Least Squares n Functon Approxmaton...
More information1/10/18. Definitions. Probabilistic models. Why probabilistic models. Example: a fair 6-sided dice. Probability
/0/8 I529: Machne Learnng n Bonformatcs Defntons Probablstc models Probablstc models A model means a system that smulates the object under consderaton A probablstc model s one that produces dfferent outcomes
More informationis the calculated value of the dependent variable at point i. The best parameters have values that minimize the squares of the errors
Multple Lnear and Polynomal Regresson wth Statstcal Analyss Gven a set of data of measured (or observed) values of a dependent varable: y versus n ndependent varables x 1, x, x n, multple lnear regresson
More information3.1 Expectation of Functions of Several Random Variables. )' be a k-dimensional discrete or continuous random vector, with joint PMF p (, E X E X1 E X
Statstcs 1: Probablty Theory II 37 3 EPECTATION OF SEVERAL RANDOM VARIABLES As n Probablty Theory I, the nterest n most stuatons les not on the actual dstrbuton of a random vector, but rather on a number
More information9 : Learning Partially Observed GM : EM Algorithm
10-708: Probablstc Graphcal Models 10-708, Sprng 2012 9 : Learnng Partally Observed GM : EM Algorthm Lecturer: Erc P. Xng Scrbes: Mrnmaya Sachan, Phan Gadde, Vswanathan Srpradha 1 Introducton So far n
More informationModeling and Simulation NETW 707
Modelng and Smulaton NETW 707 Lecture 5 Tests for Random Numbers Course Instructor: Dr.-Ing. Magge Mashaly magge.ezzat@guc.edu.eg C3.220 1 Propertes of Random Numbers Random Number Generators (RNGs) must
More information