Information Retrieval Language models for IR
|
|
- Robert Fleming
- 5 years ago
- Views:
Transcription
1 Informaton Retreval Language models for IR From Mannng and Raghavan s course [Borros sldes from Vktor Lavrenko and Chengxang Zha] 1
2 Recap Tradtonal models Boolean model Vector space model robablstc models Today IR usng statstcal language models 2
3 rncple of statstcal language modelng Goal: create a statstcal model so that one can calculate the probablty of a seuence of ords s 1, 2,, n n a language. General approach: Tranng corpus s robabltes of the observed elements s 3
4 Examples of utlzaton Speech recognton Tranng corpus sgnals + ords probabltes: ordsgnal, ord2ord1 Utlzaton: sgnals seuence of ords Statstcal taggng Tranng corpus ords + tags n, v robabltes: ordtag, tag2tag1 Utlzaton: sentence seuence of tags 4
5 Stochastc Language Models A statstcal model for generatng text robablty dstrbuton over strngs n a gven language M M M M, M, M, 5
6 6 rob. of a seuence of ords,... 2, 1 n s Elements to be estmated: - If h s too long, one cannot observe h, n the tranng corpus, and h, s hard generalze - Soluton: lmt the length of h h h h n n n h 1 1 1,
7 7 n-grams Lmt h to n-1 precedng ords Most used cases Un-gram: B-gram: Tr-gram: n s 1 n s 1 1 n s 1 1 2
8 Ungram and hgher-order models Ungram Language Models Bgram generally, n-gram Language Models Easy. Effectve! 8
9 Estmaton Hstory: short long modelng: coarse refned Estmaton: easy dffcult Maxmum lkelhood estmaton MLE # # h h C C un If h m s not observed n tranng corpus, h0 h m coud stll be possble n the language Soluton: smoothng ngram 9
10 Smoothng Goal: assgn a lo probablty to ords or n- grams not observed n the tranng corpus MLE smoothed ord 10
11 Smoothng methods n-gram: α Change the fre. of occurrences Laplace smoothng add-one: Good-Turng change the fre. r to r* add _ one nr no. of n-grams of fre. r α C α V r* r + 1 α + 1 α + 1 redstrbute the total count of ords of freuency r+1 to ords of freuency r n r+ 1 n r 11
12 12 Smoothng cont d Combne a model th a loer-order model Backoff Katz Interpolaton Jelnek-Mercer In IR, combne doc. th corpus otherse 0 f > Katz GT Katz α JM ML JM + λ λ 1 C D D ML ML λ λ +
13 Standard robablstc IR Informaton need d1 uery matchng d2 dn document collecton 13
14 IR based on Language Model LM Informaton need Q M 1 d M d d1 uery generaton M d2 d2 A uery generaton process For an nformaton need, magne an deal document Imagne hat ords could appear n that document Formulate a uery usng those ords M dn dn document collecton 14
15 Stochastc Language Models Models probablty of generatng strngs n the language commonly all strngs over alphabet Model M 0.2 the 0.1 a 0.01 man 0.01 oman 0.03 sad 0.02 lkes the man lkes the oman multply s M
16 Stochastc Language Models Model probablty of generatng any strng Model M1 Model M2 0.2 the 0.01 class sayst pleaseth yon maden 0.01 oman 0.2 the class 0.03 sayst 0.02 pleaseth 0.1 yon 0.01 maden oman the class pleaseth yon maden sm2 > sm1 16
17 Usng Language Models n IR Treat each document as the bass for a model e.g., ungram suffcent statstcs Rank document d based on d d d x d / s the same for all documents, so gnore d [the pror] s often treated as the same for all d But e could use crtera lke authorty, length, genre d s the probablty of gven d s model Very general formal approach 17
18 Language Models for IR Language Modelng Approaches Attempt to model uery generaton process Documents are ranked by the probablty that a uery ould be observed as a random sample from the respectve document model Multnomal approach 18
19 Retreval based on probablstc LM Treat the generaton of ueres as a random process. Approach Infer a language model for each document. Estmate the probablty of generatng the uery accordng to each of these models. Rank the documents accordng to these probabltes. Usually a ungram estmate of ords s used 19
20 Retreval based on probablstc LM Intuton Users Have a reasonable dea of terms that are lkely to occur n documents of nterest. They ll choose uery terms that dstngush these documents from others n the collecton. Collecton statstcs Are ntegral parts of the language model. Are not used heurstcally as n many other approaches. In theory. In practce, there s usually some ggle room for emprcally set parameters 20
21 Query generaton probablty 1 Rankng formula p Q, d The probablty of producng the uery gven the language model of document d usng MLE s: pˆ Q M pˆ t M d t Q tf dl t Q t, d d p d p Q p d p Q ml d d M d Ungram assumpton: Gven a partcular language model, the uery terms occur ndependently tf M d t, d dl d : language model of document d : ra tf of term t n document d : total number of tokens n document d 21
22 Insuffcent data Zero probablty May not sh to assgn a probablty of zero to a document that s mssng one or more of the uery terms [gves conjuncton semantcs] General approach p t M d A non-occurrng term s possble, but no more lkely than ould be expected by chance n the collecton. If tf 0, t, d cs cf t : ra collecton szetotal number of tokens n the collecton : ra count of term t n the collecton 0 22
23 Insuffcent data Zero probabltes spell dsaster We need to smooth probabltes Dscount nonzero probabltes Gve some probablty mass to unseen thngs There s a de space of approaches to smoothng probablty dstrbutons to deal th ths problem, such as addng 1, ½ or ε to counts, Drchlet prors, dscountng, and nterpolaton A smple dea that orks ell n practce s to use a mxture beteen the document multnomal and the collecton multnomal dstrbuton 23
24 Mxture model d λ mle M d + 1 λ mle M c Mxes the probablty from the document th the general collecton freuency of the ord. Correctly settng λ s very mportant A hgh value of lambda makes the search conjunctve-lke sutable for short ueres A lo value s more sutable for long ueres Can tune λ to optmze performance erhaps make t dependent on document sze cf. Drchlet pror or Wtten-Bell smoothng 24
25 Basc mxture model summary General formulaton of the LM for IR p Q, d p d 1 λ p t + λp t t Q M d general language model ndvdual-document model The user has a document n mnd, and generates the uery from ths document. The euaton represents the probablty that the document that the user had n mnd as n fact ths one. 25
26 Example Document collecton 2 documents d 1 : Xerox reports a proft but revenue s don d 2 : Lucent narros uarter loss but revenue decreases further Model: MLE ungram from documents; λ ½ Query: revenue don Qd 1 [1/8 + 2/16/2] x [1/8 + 1/16/2] 1/8 x 3/32 3/256 Qd 2 [1/8 + 2/16/2] x [0 + 1/16/2] 1/8 x 1/32 1/256 Rankng: d 1 > d 2 26
27 onte and Croft Experments Data TREC topcs on TREC dsks 2 and 3 Natural language ueres consstng of one sentence each TREC topcs on TREC dsk 3 usng the concept felds Lsts of good terms <num>number: 054 <dom>doman: Internatonal Economcs <ttle>topc: Satellte Launch Contracts <desc>descrpton: </desc> <con>concepts: 1. Contract, agreement 2. Launch vehcle, rocket, payload, satellte Launch servces, </con>
28 recson/recall results
29 recson/recall results
30 LM vs. rob. Model for IR The man dfference s hether Relevance fgures explctly n the model or not LM approach attempts to do aay th modelng relevance LM approach asssumes that documents and expressons of nformaton problems are of the same type Computatonally tractable, ntutvely appealng 30
31 LM vs. rob. Model for IR roblems of basc LM approach Assumpton of euvalence beteen document and nformaton problem representaton s unrealstc Very smple models of language Can t easly accommodate phrases, passages, Boolean operators Several extensons puttng relevance back nto the model, uery expanson term dependences, etc. 31
32 Alternatve Models of Text Generaton M Searcher Query M Searcher Query Model Query Is ths the same model? Wrter M Wrter Doc Model Doc M Doc 32
33 Model Comparson Estmate uery and document models and compare Sutable measure s KL dvergence DQ m D m euvalent to uery-lkelhood approach f smple emprcal dstrbuton used for uery model hy? More general rsk mnmzaton frameork has been proposed Zha and Lafferty 2001 Better results than uery-lkelhood 33
34 Comparson Wth Vector Space There s some relaton to tradtonal tf.df models: unscaled term freuency s drectly n model the probabltes do length normalzaton of term freuences the effect of dong a mxture th overall collecton freuences s a lttle lke df: terms rare n the general collecton but common n some documents ll have a greater nfluence on the rankng 34
35 35 Comparason: LM v.s. tf*df 1,, 1, 1, 1,, 1 /, 1, 1,, 1, 1, D Q D Q j D Q j Q D Q j D Q D Q j j j C C tf D D tf const C C tf C C tf D D tf const C C tf C C tf C C tf D D tf C C tf C C tf D D tf D D Q λ λ λ λ λ λ λ λ λ λ λ λ Log QD ~ VSM th tf*df and document length normalzaton Smoothng ~ df + length normalzaton df
36 Comparson Wth Vector Space Smlar n some ays Term eghts based on freuency Terms often used as f they ere ndependent Inverse document/collecton freuency used Some form of length normalzaton useful Dfferent n others Based on probablty rather than smlarty Intutons are probablstc rather than geometrc Detals of use of document length and term, document, and collecton freuency dffer 36
37 Resources J.M. onte and W.B. Croft A language modelng approach to nformaton retreval. In SIGIR 21. D. Hemstra A lngustcally motvated probablstc model of nformaton retreval. ECDL 2, pp A. Berger and J. Lafferty Informaton retreval as statstcal translaton. SIGIR 22, pp D.R.H. Mller, T. Leek, and R.M. Schartz A hdden Markov model nformaton retreval system. SIGIR 22, pp Chengxang Zha, Statstcal language models for nformaton retreval, n the seres of Synthess Lectures on Human Language Technologes, Morgan & Claypool, 2009 [Several relevant neer papers at SIGIR 2000 no.] Workshop on Language Modelng and Informaton Retreval, CMU The Lemur Toolkt for Language Modelng and Informaton Retreval. CMU/Umass LM and IR system n C++, currently actvely developed. 37
Retrieval Models: Language models
CS-590I Informaton Retreval Retreval Models: Language models Luo S Department of Computer Scence Purdue Unversty Introducton to language model Ungram language model Document language model estmaton Maxmum
More informationCS47300: Web Information Search and Management
CS47300: Web Informaton Search and Management Probablstc Retreval Models Prof. Chrs Clfton 7 September 2018 Materal adapted from course created by Dr. Luo S, now leadng Albaba research group 14 Why probabltes
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 informationProbabilistic Information Retrieval CE-324: Modern Information Retrieval Sharif University of Technology
Probablstc Informaton Retreval CE-324: Modern Informaton Retreval Sharf Unversty of Technology M. Soleyman Fall 2016 Most sldes have been adapted from: Profs. Mannng, Nayak & Raghavan (CS-276, Stanford)
More informationQuestion Classification Using Language Modeling
Queston Classfcaton Usng Language Modelng We L Center for Intellgent Informaton Retreval Department of Computer Scence Unversty of Massachusetts, Amherst, MA 01003 ABSTRACT Queston classfcaton assgns a
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 informationWeb-Mining Agents Probabilistic Information Retrieval
Web-Mnng Agents Probablstc Informaton etreval Prof. Dr. alf Möller Unverstät zu Lübeck Insttut für Informatonssysteme Karsten Martny Übungen Acknowledgements Sldes taken from: Introducton to Informaton
More informationExtending Relevance Model for Relevance Feedback
Extendng Relevance Model for Relevance Feedback Le Zhao, Chenmn Lang and Jame Callan Language Technologes Insttute School of Computer Scence Carnege Mellon Unversty {lezhao, chenmnl, callan}@cs.cmu.edu
More informationExpectation 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 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 informationYou have an information need to monitor, say:
TEXT CLASSIFICATION STANDING QUERIES The ath from IR to text classfcaton: You have an nformaton need to montor, say: Unrest n the Nger delta regon You want to rerun an arorate query erodcally to fnd new
More informationCorpora and Statistical Methods Lecture 6. Semantic similarity, vector space models and wordsense disambiguation
Corpora and Statstcal Methods Lecture 6 Semantc smlarty, vector space models and wordsense dsambguaton Part 1 Semantc smlarty Synonymy Dfferent phonologcal/orthographc words hghly related meanngs: sofa
More informationNote on EM-training of IBM-model 1
Note on EM-tranng of IBM-model INF58 Language Technologcal Applcatons, Fall The sldes on ths subject (nf58 6.pdf) ncludng the example seem nsuffcent to gve a good grasp of what s gong on. Hence here are
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 informationComputation of Higher Order Moments from Two Multinomial Overdispersion Likelihood Models
Computaton of Hgher Order Moments from Two Multnomal Overdsperson Lkelhood Models BY J. T. NEWCOMER, N. K. NEERCHAL Department of Mathematcs and Statstcs, Unversty of Maryland, Baltmore County, Baltmore,
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 informationSupport Vector Machines CS434
Support Vector Machnes CS434 Lnear Separators Many lnear separators exst that perfectly classfy all tranng examples Whch of the lnear separators s the best? Intuton of Margn Consder ponts A, B, and C We
More informationRanked Retrieval (2)
Text Technologies for Data Science INFR11145 Ranked Retrieval (2) Instructor: Walid Magdy 31-Oct-2017 Lecture Objectives Learn about Probabilistic models BM25 Learn about LM for IR 2 1 Recall: VSM & TFIDF
More informationComputing MLE Bias Empirically
Computng MLE Bas Emprcally Kar Wa Lm Australan atonal Unversty January 3, 27 Abstract Ths note studes the bas arses from the MLE estmate of the rate parameter and the mean parameter of an exponental dstrbuton.
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 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 informationStat260: Bayesian Modeling and Inference Lecture Date: February 22, Reference Priors
Stat60: Bayesan Modelng and Inference Lecture Date: February, 00 Reference Prors Lecturer: Mchael I. Jordan Scrbe: Steven Troxler and Wayne Lee In ths lecture, we assume that θ R; n hgher-dmensons, reference
More informationHidden Markov Models
Note to other teachers and users of these sldes. Andrew would be delghted f you found ths source materal useful n gvng your own lectures. Feel free to use these sldes verbatm, or to modfy them to ft your
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 informationLecture 6 More on Complete Randomized Block Design (RBD)
Lecture 6 More on Complete Randomzed Block Desgn (RBD) Multple test Multple test The multple comparsons or multple testng problem occurs when one consders a set of statstcal nferences smultaneously. For
More information9/10/18. CS 6120/CS4120: Natural Language Processing. Outline. Probabilistic Language Models. Probabilistic Language Modeling
Outlne CS 6120/CS4120: Natural Language Processng Instructor: Prof. Lu Wang College of Computer and Informaton Scence Northeastern Unversty Webpage: www.ccs.neu.edu/home/luwang Probablstc language model
More information4.1. Lecture 4: Fitting distributions: goodness of fit. Goodness of fit: the underlying principle
Lecture 4: Fttng dstrbutons: goodness of ft Goodness of ft Testng goodness of ft Testng normalty An mportant note on testng normalty! L4.1 Goodness of ft measures the extent to whch some emprcal dstrbuton
More informationMSU at ImageCLEF: Cross Language and Interactive Image Retrieval
MSU at ImageCLEF: Cross Language and Interactve Image Retreval Vneet Bansal, Chen Zhang, Joyce Y. Cha, Rong Jn Department of Computer Scence and Engneerng, Mchgan State Unversty East Lansng, MI48824, U.S.A.
More informationProbabilistic Ranking Principle. Hongning Wang
robablstc ankng rncple Hongnng Wang CS@UVa Noton of relevance elevance epq epd Smlarty r1qd r {01} robablty of elevance d q or q d robablstc nference Dfferent rep & smlarty Vector space model Salton et
More informationConjugacy and the Exponential Family
CS281B/Stat241B: Advanced Topcs n Learnng & Decson Makng Conjugacy and the Exponental Famly Lecturer: Mchael I. Jordan Scrbes: Bran Mlch 1 Conjugacy In the prevous lecture, we saw conjugate prors for the
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 informationLearning from Data 1 Naive Bayes
Learnng from Data 1 Nave Bayes Davd Barber dbarber@anc.ed.ac.uk course page : http://anc.ed.ac.uk/ dbarber/lfd1/lfd1.html c Davd Barber 2001, 2002 1 Learnng from Data 1 : c Davd Barber 2001,2002 2 1 Why
More informationPredictive Analytics : QM901.1x Prof U Dinesh Kumar, IIMB. All Rights Reserved, Indian Institute of Management Bangalore
Sesson Outlne Introducton to classfcaton problems and dscrete choce models. Introducton to Logstcs Regresson. Logstc functon and Logt functon. Maxmum Lkelhood Estmator (MLE) for estmaton of LR parameters.
More informationMaxent Models & Deep Learning
Maxent Models & Deep Learnng 1. Last bts of maxent (sequence) models 1.MEMMs vs. CRFs 2.Smoothng/regularzaton n maxent models 2. Deep Learnng 1. What s t? Why s t good? (Part 1) 2. From logstc regresson
More informationMidterm Review. Hongning Wang
Mdterm Revew Hongnng Wang CS@UVa Core concepts Search Engne Archtecture Key components n a modern search engne Crawlng & Text processng Dfferent strateges for crawlng Challenges n crawlng Text processng
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 informationA Probabilistic Multimedia Retrieval Model and its Evaluation
A Probablstc Multmeda Retreval Model and ts Evaluaton Thjs Westerveld 1, Arjen P. de Vres 1, Alex van Ballegooj 1, Francska de Jong 2, and Djoerd Hemstra 2 1 CWI PO Box 94079, 1090 GB Amsterdam The Netherlands
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 informationProbabilistic Ranking Principle. Hongning Wang
Probablstc anng Prncple Hongnng Wang CS@UVa ecap: latent semantc analyss Soluton Low ran matrx approxmaton Imagne ths s *true* concept-document matrx Imagne ths s our observed term-document matrx andom
More informationChecking Pairwise Relationships. Lecture 19 Biostatistics 666
Checkng Parwse Relatonshps Lecture 19 Bostatstcs 666 Last Lecture: Markov Model for Multpont Analyss X X X 1 3 X M P X 1 I P X I P X 3 I P X M I 1 3 M I 1 I I 3 I M P I I P I 3 I P... 1 IBD states along
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 informationMore metrics on cartesian products
More metrcs on cartesan products If (X, d ) are metrc spaces for 1 n, then n Secton II4 of the lecture notes we defned three metrcs on X whose underlyng topologes are the product topology The purpose of
More informationNumerical Heat and Mass Transfer
Master degree n Mechancal Engneerng Numercal Heat and Mass Transfer 06-Fnte-Dfference Method (One-dmensonal, steady state heat conducton) Fausto Arpno f.arpno@uncas.t Introducton Why we use models and
More informationSee Book Chapter 11 2 nd Edition (Chapter 10 1 st Edition)
Count Data Models See Book Chapter 11 2 nd Edton (Chapter 10 1 st Edton) Count data consst of non-negatve nteger values Examples: number of drver route changes per week, the number of trp departure changes
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 informationBenchmarking in pig production
Benchmarkng n pg producton Thomas Algot Søllested Egeberg Internatonal A/S Agenda Who am I? Benchmarkng usng Data Envelopment Analyss Focus-Fnder an example of benchmarkng n pg producton 1 Who am I? M.Sc.
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 informationDirichlet Mixtures in Text Modeling
Drchlet Mxtures n Text Modelng Mko Yamamoto and Kugatsu Sadamtsu CS Techncal report CS-TR-05-1 Unversty of Tsukuba May 30, 2005 Abstract Word rates n text vary accordng to global factors such as genre,
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 informationSimulated Power of the Discrete Cramér-von Mises Goodness-of-Fit Tests
Smulated of the Cramér-von Mses Goodness-of-Ft Tests Steele, M., Chaselng, J. and 3 Hurst, C. School of Mathematcal and Physcal Scences, James Cook Unversty, Australan School of Envronmental Studes, Grffth
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 informationConvergence of random processes
DS-GA 12 Lecture notes 6 Fall 216 Convergence of random processes 1 Introducton In these notes we study convergence of dscrete random processes. Ths allows to characterze phenomena such as the law of large
More informationLecture 6 Hidden Markov Models and Maximum Entropy Models
Lecture 6 Hdden Markov Models and Maxmum Entropy Models CS 6320 82 HMM Outlne Markov Chans Hdden Markov Model Lkelhood: Forard Alg. Decodng: Vterb Alg. Maxmum Entropy Models 83 Dentons A eghted nte-state
More informationSelecting Good Expansion Terms for Pseudo-Relevance Feedback
Guhong Cao, Jan-Yun Ne Department of Computer Scence and Operatons Research Unversty of Montreal, Canada {caogu, ne}@ro.umontreal.ca Selectng Good Expanson erms for Pseudo-Relevance Feedback ABSRAC Pseudo-relevance
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 informationMaximum Likelihood Estimation (MLE)
Maxmum Lkelhood Estmaton (MLE) Ken Kreutz-Delgado (Nuno Vasconcelos) ECE 175A Wnter 01 UCSD Statstcal Learnng Goal: Gven a relatonshp between a feature vector x and a vector y, and d data samples (x,y
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 informationThis excerpt from. Foundations of Statistical Natural Language Processing. Christopher D. Manning and Hinrich Schütze The MIT Press.
Ths excerpt from Foundatons of Statstcal Natural Language Processng. Chrstopher D. Mannng and Hnrch Schütze. 1999 The MIT Press. s provded n screen-vewable form for personal use only by members of MIT
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 informationMulti-Conditional Learning for Joint Probability Models with Latent Variables
Mult-Condtonal Learnng for Jont Probablty Models wth Latent Varables Chrs Pal, Xueru Wang, Mchael Kelm and Andrew McCallum Department of Computer Scence Unversty of Massachusetts Amherst Amherst, MA USA
More informationU-Pb Geochronology Practical: Background
U-Pb Geochronology Practcal: Background Basc Concepts: accuracy: measure of the dfference between an expermental measurement and the true value precson: measure of the reproducblty of the expermental result
More information1 Binary Response Models
Bnary and Ordered Multnomal Response Models Dscrete qualtatve response models deal wth dscrete dependent varables. bnary: yes/no, partcpaton/non-partcpaton lnear probablty model LPM, probt or logt models
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 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 informationNUMERICAL DIFFERENTIATION
NUMERICAL DIFFERENTIATION 1 Introducton Dfferentaton s a method to compute the rate at whch a dependent output y changes wth respect to the change n the ndependent nput x. Ths rate of change s called the
More informationChapter 13: Multiple Regression
Chapter 13: Multple Regresson 13.1 Developng the multple-regresson Model The general model can be descrbed as: It smplfes for two ndependent varables: The sample ft parameter b 0, b 1, and b are used to
More information15-381: Artificial Intelligence. Regression and cross validation
15-381: Artfcal Intellgence Regresson and cross valdaton Where e are Inputs Densty Estmator Probablty Inputs Classfer Predct category Inputs Regressor Predct real no. Today Lnear regresson Gven an nput
More informationClustering & Unsupervised Learning
Clusterng & Unsupervsed Learnng Ken Kreutz-Delgado (Nuno Vasconcelos) ECE 175A Wnter 2012 UCSD Statstcal Learnng Goal: Gven a relatonshp between a feature vector x and a vector y, and d data samples (x,y
More informationHidden Markov Models & The Multivariate Gaussian (10/26/04)
CS281A/Stat241A: Statstcal Learnng Theory Hdden Markov Models & The Multvarate Gaussan (10/26/04) Lecturer: Mchael I. Jordan Scrbes: Jonathan W. Hu 1 Hdden Markov Models As a bref revew, hdden Markov models
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 informationANSWERS. Problem 1. and the moment generating function (mgf) by. defined for any real t. Use this to show that E( U) var( U)
Econ 413 Exam 13 H ANSWERS Settet er nndelt 9 deloppgaver, A,B,C, som alle anbefales å telle lkt for å gøre det ltt lettere å stå. Svar er gtt . Unfortunately, there s a prntng error n the hnt of
More informationLanguage Modeling. Mausam. (Based on slides of Michael Collins, Dan Jurafsky, Dan Klein, Chris Manning, Luke Zettlemoyer)
Language Modelng Mausam Based on sldes of Mchael Collns, Dan Jurafsky, Dan Klen, Chrs Mannng, Luke Zettlemoyer Outlne Motvaton Task Defnton Probablty Estmaton Evaluaton Smoothng 2 Smple Interpolaton and
More informationANOVA. The Observations y ij
ANOVA Stands for ANalyss Of VArance But t s a test of dfferences n means The dea: The Observatons y j Treatment group = 1 = 2 = k y 11 y 21 y k,1 y 12 y 22 y k,2 y 1, n1 y 2, n2 y k, nk means: m 1 m 2
More information= z 20 z n. (k 20) + 4 z k = 4
Problem Set #7 solutons 7.2.. (a Fnd the coeffcent of z k n (z + z 5 + z 6 + z 7 + 5, k 20. We use the known seres expanson ( n+l ( z l l z n below: (z + z 5 + z 6 + z 7 + 5 (z 5 ( + z + z 2 + z + 5 5
More information4 Analysis of Variance (ANOVA) 5 ANOVA. 5.1 Introduction. 5.2 Fixed Effects ANOVA
4 Analyss of Varance (ANOVA) 5 ANOVA 51 Introducton ANOVA ANOVA s a way to estmate and test the means of multple populatons We wll start wth one-way ANOVA If the populatons ncluded n the study are selected
More informationTopic 23 - Randomized Complete Block Designs (RCBD)
Topc 3 ANOVA (III) 3-1 Topc 3 - Randomzed Complete Block Desgns (RCBD) Defn: A Randomzed Complete Block Desgn s a varant of the completely randomzed desgn (CRD) that we recently learned. In ths desgn,
More informationContinuous vs. Discrete Goods
CE 651 Transportaton Economcs Charsma Choudhury Lecture 3-4 Analyss of Demand Contnuous vs. Dscrete Goods Contnuous Goods Dscrete Goods x auto 1 Indfference u curves 3 u u 1 x 1 0 1 bus Outlne Data Modelng
More informationSTAT 3008 Applied Regression Analysis
STAT 3008 Appled Regresson Analyss Tutoral : Smple Lnear Regresson LAI Chun He Department of Statstcs, The Chnese Unversty of Hong Kong 1 Model Assumpton To quantfy the relatonshp between two factors,
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 informationBOOTSTRAP METHOD FOR TESTING OF EQUALITY OF SEVERAL MEANS. M. Krishna Reddy, B. Naveen Kumar and Y. Ramu
BOOTSTRAP METHOD FOR TESTING OF EQUALITY OF SEVERAL MEANS M. Krshna Reddy, B. Naveen Kumar and Y. Ramu Department of Statstcs, Osmana Unversty, Hyderabad -500 007, Inda. nanbyrozu@gmal.com, ramu0@gmal.com
More informationHow its computed. y outcome data λ parameters hyperparameters. where P denotes the Laplace approximation. k i k k. Andrew B Lawson 2013
Andrew Lawson MUSC INLA INLA s a relatvely new tool that can be used to approxmate posteror dstrbutons n Bayesan models INLA stands for ntegrated Nested Laplace Approxmaton The approxmaton has been known
More informationAnother Language Example. More on NLP. Statistical NLP at Work. Statistical NLP. Lexical Terms. Lexical Resources
More on LP An ecerpt from Games magazne, ovember 2005 The only smurf we have to smurf s smurf tself. Yea, though I smurf through the smurf of the smurf of smurf, I wll smurf no smurf. The frst smurf about
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 informationx = , so that calculated
Stat 4, secton Sngle Factor ANOVA notes by Tm Plachowsk n chapter 8 we conducted hypothess tests n whch we compared a sngle sample s mean or proporton to some hypotheszed value Chapter 9 expanded ths to
More information2E Pattern Recognition Solutions to Introduction to Pattern Recognition, Chapter 2: Bayesian pattern classification
E395 - Pattern Recognton Solutons to Introducton to Pattern Recognton, Chapter : Bayesan pattern classfcaton Preface Ths document s a soluton manual for selected exercses from Introducton to Pattern Recognton
More informationLecture 12: Classification
Lecture : Classfcaton g Dscrmnant functons g The optmal Bayes classfer g Quadratc classfers g Eucldean and Mahalanobs metrcs g K Nearest Neghbor Classfers Intellgent Sensor Systems Rcardo Guterrez-Osuna
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 informationComparison of Regression Lines
STATGRAPHICS Rev. 9/13/2013 Comparson of Regresson Lnes Summary... 1 Data Input... 3 Analyss Summary... 4 Plot of Ftted Model... 6 Condtonal Sums of Squares... 6 Analyss Optons... 7 Forecasts... 8 Confdence
More informationStat 642, Lecture notes for 01/27/ d i = 1 t. n i t nj. n j
Stat 642, Lecture notes for 01/27/05 18 Rate Standardzaton Contnued: Note that f T n t where T s the cumulatve follow-up tme and n s the number of subjects at rsk at the mdpont or nterval, and d s the
More informationA PROBABILITY-DRIVEN SEARCH ALGORITHM FOR SOLVING MULTI-OBJECTIVE OPTIMIZATION PROBLEMS
HCMC Unversty of Pedagogy Thong Nguyen Huu et al. A PROBABILITY-DRIVEN SEARCH ALGORITHM FOR SOLVING MULTI-OBJECTIVE OPTIMIZATION PROBLEMS Thong Nguyen Huu and Hao Tran Van Department of mathematcs-nformaton,
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 informationTwo-factor model. Statistical Models. Least Squares estimation in LM two-factor model. Rats
tatstcal Models Lecture nalyss of Varance wo-factor model Overall mean Man effect of factor at level Man effect of factor at level Y µ + α + β + γ + ε Eε f (, ( l, Cov( ε, ε ) lmr f (, nteracton effect
More informationMASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 12 10/21/2013. Martingale Concentration Inequalities and Applications
MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.65/15.070J Fall 013 Lecture 1 10/1/013 Martngale Concentraton Inequaltes and Applcatons Content. 1. Exponental concentraton for martngales wth bounded ncrements.
More informationCross-Language Information Retrieval (CLIR)
Cross-Language Informaton Retreval CLIR Ananthakrshnan R Computer Scence & Engg. IIT Bombay anand@cse Aprl 7 2006 Natural Language Processng/Language Technology for the Web Cross Language Informaton Retreval
More informationDigital Modems. Lecture 2
Dgtal Modems Lecture Revew We have shown that both Bayes and eyman/pearson crtera are based on the Lkelhood Rato Test (LRT) Λ ( r ) < > η Λ r s called observaton transformaton or suffcent statstc The crtera
More informationSolutions Homework 4 March 5, 2018
1 Solutons Homework 4 March 5, 018 Soluton to Exercse 5.1.8: Let a IR be a translaton and c > 0 be a re-scalng. ˆb1 (cx + a) cx n + a (cx 1 + a) c x n x 1 cˆb 1 (x), whch shows ˆb 1 s locaton nvarant and
More informationEcon107 Applied Econometrics Topic 9: Heteroskedasticity (Studenmund, Chapter 10)
I. Defnton and Problems Econ7 Appled Econometrcs Topc 9: Heteroskedastcty (Studenmund, Chapter ) We now relax another classcal assumpton. Ths s a problem that arses often wth cross sectons of ndvduals,
More information10.34 Fall 2015 Metropolis Monte Carlo Algorithm
10.34 Fall 2015 Metropols Monte Carlo Algorthm The Metropols Monte Carlo method s very useful for calculatng manydmensonal ntegraton. For e.g. n statstcal mechancs n order to calculate the prospertes of
More informationVapnik-Chervonenkis theory
Vapnk-Chervonenks theory Rs Kondor June 13, 2008 For the purposes of ths lecture, we restrct ourselves to the bnary supervsed batch learnng settng. We assume that we have an nput space X, and an unknown
More informationIntroduction to Vapor/Liquid Equilibrium, part 2. Raoult s Law:
CE304, Sprng 2004 Lecture 4 Introducton to Vapor/Lqud Equlbrum, part 2 Raoult s Law: The smplest model that allows us do VLE calculatons s obtaned when we assume that the vapor phase s an deal gas, and
More information