If you have a spare hour and a half, I highly recommend you watch Percy Liang’s entire talk which this summary article was based on: Special thanks to Melissa Fabros for recommending Percy’s talk, Matthew Kleinsmith for highlighting the MIT Media Lab definition of “grounded” language, and Jeremy Howard and Rachel Thomas of fast.ai for faciliating our connection and conversation. /Length 1337 [, Mon 10/22: Lecture 9: VC dimension, covering techniques pliang@cs.stanford.edu. offerings of this course, Peter Bartlett's statistical learning theory course, Boyd and A few pointers: Our simple example came from this nice article by Percy Liang. Percy Liang Department of Computer Science Stanford University Stanford, CA 94305 Abstract In user-facing applications, displaying calibrated confidence measures— probabilities that correspond to true frequency—can be as important as obtaining high accuracy. /Filter /FlateDecode OpenURL . Percy Liang's course notes from previous offerings of this course. x���o�6���+t��Z��.CV��=�;02c���#M�חI�q�6Z���N�h�����%-#�y��6��5d�)��D��H�qq�SL�"��. Peter Bartlett's statistical learning theory course. Wassersetin GANs Percy Liang This short note presents a new formal language, lambda dependency-based compositional semantics (lambda DCS) for representing logical forms in semantic parsing. Thompson Sampling [, Mon 12/03: Lecture 19: Regret bound for UCB, Bayesian setup, Boyd and Vandenberghe's Convex Optimization. This preview shows page 1 - 3 out of 12 pages. CS229T/STAT231: Statistical Learning Theory (Winter 2016) Percy Liang Last updated Wed Apr 20 2016 01:36 These lecture notes will be updated periodically as the course goes on. online learning EMNLP 2019 (long papers). �8YX�.��?��,�8�#���C@%�)�, �XWd��A@ɔ�����B\J�b\��3�/P�p�Q��(���I�ABAe�h��%���o�5�����[u��~���������x���C�~yo;Z����@�o��o�#����'�:� �u$��'���4ܕMWw~fmW��V~]�%�@��U+7F�`�r������@�!�U�+G��m��I�a��,]����Ҳ�,�!��}���.�-��4H����+Wu����/��Z9�3qno}ٗ��n�i}��M�f��l[T���K B�Qa;�Onl���e����`�$~���o]N���". Liang, a senior majoring in computer science and minoring in music and also a student in the Master of Engineering program, will present an Advanced Music Performance piano recital today (March 17) at 5 p.m. in Killian Hall. Approximation in Shallow NN. linear algebra, We scraped Piazza question, answers, tags, followups, and notes from the Autumn 2016 offering of CS 221 as well as the 2013 - 2016 offerings of CS 124, with the permission of Professors Percy Liang and Dan Jurafsky, respectively. In particular, I am interested in executable representations such as database queries or … … Percy Liang ; Roweis and Saul ; Percy Liang ; Amos Storkey ; PCA : M. Girolami ; Andrew Ng ; Kevin Murphy ; Amos Storkey ; Lindsay Smith ; Kevin Murphy ; Model Selection: Topic Notes Slides Reading Homework; Model Selection/Comparison : Andrew Ng ; Zoubin Ghahramani ; Parameter estimation/Optimization techniques Topic Notes Slides Reading Homework; Parameter estimation : … Fp(t�� ��%4@@G���q�\ To scale up influence functions to modern machine learning … Better bound? Abstract. [, Mon 10/15: Lecture 7: Rademacher complexity, neural networks Percy Liang. 2 0 obj << %���� Better basis? /N 100 Percy Liang Computer Forum April 16, 2013 ... Summary so far: Modeling deep semantics of natural language is important Need to learn from natural/weak supervision to obtain broad coverage Rest of talk: Spectral methods for learning latentvariable models Learning a broad coverage semantic parser 11. EM: Revisiting K-Means 53 1Reference: Percy Liang, CS221 (2015) • EM tries to maximize marginal likelihood • K-means • Is a special case of EM (for GMMs with variance tending to 0) • Objective: Estimate cluster centers • But don’t know which points belong to which clusters • Take an alternating optimization approach • Find the … How does it improve bound for various classes of functions? Derivation for linear regression. Stanford University. stream OpenURL . [, Wed 10/24: Lecture 10: Covering techniques, overview of GANs [, Thu 11/01: Homework 2 (uniform convergence), Mon 11/05: Lecture 13: Restricted Approximability, overview of stream Certificate. >> two-layer neural networks … Amita Kamath Robin Jia Percy Liang Computer Science Department, Stanford University fkamatha, robinjia, pliangg@cs.stanford.edu Abstract To avoid giving wrong answers, question an-swering (QA) models need to know when to abstain from answering. probability theory, Martin Wainwright's statistical learning theory course /First 813 According to media reports, a pair of hackers said on Saturday that the Firefox Web browser, commonly perceived as the safer and more customizable alternative to … percyliang has 12 repositories available. Percy Liang and Dan Klein (2007): Structured Bayesian Nonparametric Models with Variational Inference David Blei's group's topic modeling software (C, C++. Summary; Citations; Active Bibliography; Co-citation; Clustered Documents; Version History; BibTeX @MISC{Liang_learningdependency-based, author = {Percy Liang and Michael I. Jordan and Dan Klein}, title = {Learning Dependency-Based Compositional Semantics}, year = {}} Share. In order for AI to be safely deployed, the desired behavior of the AI system needs to be based on well-understood, realistic, and empirically testable assumptions. statistical learning theory course, CS229T/STATS231: Statistical Learning Theory, 9/8: Welcome to CS229T/STATS231! Statistical Learning Theory (CS229T) Lecture Notes - percyliang/cs229t >> Discriminative latent-variable models are typically learned using EM or gradient-based optimization, … In this … There is no required text for the course. /Length 1467 [, Wed 10/17: Lecture 8: Margin-based generalization error of Scribe: Percy Liang and David Malan Lecture 14: Ordered-file maintenance, analysis, order queries in lists, list labeling, external-memory model, cache-oblivious model Date: Monday, April 14, 2003 Scribe: Kunal Agrawal and Vladimir Kiriansky NAACL 2019 (short … [Please refer to, Mon 10/29: Lecture 11: Total variation distance, Wasserstein distance, Wasserstein GANs We then cleaned this data, by removing errant HTML and LaTeX symbols. endobj Pang Wei Koh 1Percy Liang Abstract How can we explain the predictions of a black-box model? Assistant Professor of Computer Science and, by courtesy, of Statistics. 378 0 obj << [, Mon 11/26: Lecture 17: Multi-armed bandit problem, general OCO with partial observation Percy Liang on Learning Hidden Computational Processes Young Kun Ko on The Hardness of Sparse PCA [pdf] Tom Griffiths on Rationality, Heuristics, and the Cost of Computation [pdf] 2.‘Statistical Learning Theory,’ Vladimir N. Vapnik, Wiley, 1998. Uploaded By sttg6. Decomposition of Errors. endstream Project: Predictable AI via Failure Detection and Robustness. and, Machine learning (CS229) or statistics (STATS315A), Convex optimization (EE364A) is recommended, Mon 09/24: Lecture 1: overview, formulation of prediction John Hewitt and Percy Liang. Deep vs. … [, Wed 11/07: Lecture 14: Online learning, online convex optimization, Follow the Leader (FTL) algorithm Percy Liang Associate Professor of Computer Science and Statistics (Courtesy) Dorsa Sadigh Assistant Professor of Computer Science and Electrical Engineering. hypothesis class [, Wed 10/03: Lecture 4: naive epsilon-cover argument, concentration inequalities Universality proof is loose: exponential number of units. The tables were randomly selected among Wikipedia tables with at least 8 rows and 5 columns. [, Wed 10/10: Lecture 6: Rademacher complexity, margin theory View Notes - 7-mdp1 from CS 221 at Stanford University. xڥW�r�6}�W�����$;�t\7�N�c��_ �0�������H'�cStg, g���]��"�IEdH�(1$""#�HĚ�RI"!��HI� �R�[���8���ʵHaQ�W�ǁl�S����}�֓����]�HF��C#�F���/K����+��֮������#�I'ꉞ�'TcϽ�G�\�7�����-��m��}�;G����6�?�paC��i\�W.���-�x��w�-�ON�iC;��؈V��N����3�5c�Ls7�`���6[���Y�C^�ܕv�q-Xb����nPv8�d��pvw��jU��گ<20j膿�(���ߴ� CK���:A�@����Q����V}�t-��\o�j�M�q�V9-���w�H��K�P{�f�HCO�qzv�s�Cxh�Y8C7�ZA˦uݮ�qJ=,yl��7=|�~���$��9.F7.�Dxz��;��G�V���8|�[˝�U�q�:G|N��G/�ӈzLb��y�������Qh�j���w�{�{ �Ptƛi�x؋TLB�S�~�Ɇx��)��N|��a�OϾ{ ��DJ�O{��`�f �|�`��j7c&aƫO�$�9{���q�C�/��]�^��t�����/���� Pages 12. A Structual Probe for Finding Syntax in Word Representations. Project Summary. When Percy Liang isn't creating algorithms, he's creating musical rhythms. statistical learning theory course, Martin Wainwright's OpenURL … Abstract Extractive reading comprehension systems can often locate the correct answer to a question in a context document, but they also tend to make unreliable guesses on questions for which the correct answer is not stated in the context. I am interested in natural language processing. Percy Liang Lots of high-dimensional data... face images Zambian President Levy Mwanawasa has won a second term in o ce in an election his challenger Michael Sata accused him of rigging, o cial results showed on Monday. Percy Liang Associate Professor of Computer Science and Statistics (courtesy) %PDF-1.5 Summary; Citations; Active Bibliography; Co-citation; Clustered Documents; Version History; BibTeX @MISC{Chaganty_spectralexperts, author = {Arun Tejasvi Chaganty and Percy Liang}, title = {Spectral Experts for Estimating Mixtures of Linear Regressions}, year = {}} Share. Notes. problems, error decomposition [, Wed 09/26: Lecture 2: asymptotics of maximum likelihood estimators (MLE) [, Mon 10/01: Lecture 3: uniform convergence overview, finite real analysis, You may also earn a Professional Certificate in … In this paper, we use influence func-tions — a classic technique from robust statis-tics — to trace a model’s prediction through the learning algorithm and back to its training data, thereby identifying training points most respon-sible for a given prediction. [, Wed 12/05: Lecture 20: Information theory, regret bound for � �T ��f��Ej͏���8���H��8f�@��)���@���D���W�a�\ ��G@Nb���� ��P� From notes of Percy Liang. Here are some areas I have worked on: Semantic parsing: Parse the input sentence into some representation of its meaning. Thompson sampling Lecture 7: MDPs I CS221: Articial Intelligence (Autumn 2013) - Percy Liang So far: search problems F B S D C E A state s, action a CS221: Pranav Rajpurkar, Robin Jia, Percy Liang. Sham Kakade's statistical learning theory course. Vandenberghe's Convex Optimization, Sham Kakade's [, Mon 11/12: Lecture 15: Follow the Regularized Leader (FTRL) algorithm /Filter /FlateDecode John Hewitt and Christopher D. Manning. [, Wed 11/28: Lecture 18: Multi-armed bandit problem in the Follow their code on GitHub. Summary; Citations; Active Bibliography; Co-citation; Clustered Documents; Version History; BibTeX @MISC{Chaganty_journalof, author = {Arun Tejasvi Chaganty and Percy Liang and C A. T. Chaganty and P. Liang and Chaganty Liang}, title = {Journal of Machine Learning Research 1–11 Supplementary Material for Spectral Experts for Estimating Mixtures of Linear Regressions}, year = {}} Share. Universality of NN. 3.‘An Elementary Introduction to Statistical Learning Theory,’ Sanjeev Kulkarni and Gilbert Harman, Wiley, 2011. (pdf) (bib) (blog) (code) (codalab) (slides) (talk). Abstract. The questions require multi-step reasoning and various data operations such as comparison, aggregation, and arithmetic computation. 1Reference: Percy Liang, CS221 (2015) 2Note: EM was first proposed in 1977. CS221: Artificial Intelligence (Autumn 2012) Percy Liang 37 Summary Linear models: prediction governed by Losscfunctions:ucapturecvarious desiderata (e.g., robustness) for both regression and binary classification (can be generalized to many other problems) Objective function: minimize loss over training data The dataset contains pairs table-question, and the respective answer. Liang, who went to high school in Arizona, has been playing piano since the age of eight and won … Amount Recommended: $255,160. [. [, Wed 11/14: Lecture 16: FTRL in concrete problems: online regression & expert problem, convex to linear reduction We are interested in calibration for structured prediction problems such as speech recognition, optical character recognition, and medical diagnosis. Existing datasets either focus exclusively on answerable questions, or use automatically generated … A number of useful references: Percy Liang's course notes from previous /Type /ObjStm Moreover, users of- ten ask questions that diverge from the model’s training data, making errors more likely and thus abstention more critical. Runner up best paper. stochastic setting [, Wed 10/31: Lecture 12: Generalization and approximation in … Compositionality: requires exponential number of units in a shallow network. Percy Liang’s Lecture Notes (Stanford) Martin Wainwright’s Lecture Notes (Berkeley) Additional References: 1.‘Learning with Kernels,’ B. Scholkopf and A. Smola, MIT Press, 2002. Designing and Interpreting Probes with Control Tasks. [, Mon 10/08: Lecture 5: Sub-Gaussian random variables, Rademacher complexity Spectral methods for learning latentvariable models (joint work with Daniel Hsu, Sham Kakade, Arun … Compositional question answering begins by mapping questions to logical forms, but training a … K�i���,% `) �Ԑ̀dR�i��t�o �l�Rl�M$Z�Ѱ��$1�)֔hXG���e*5�I��'�I��Rf2Gradgo"�4���h@E #- R x�-<>�)+��3e�M��t�`� Related. Previous years' home pages are, Uniform convergence (VC dimension, Rademacher complexity, etc), Implicit/algorithmic regularization, generalization theory for neural networks, Unsupervised learning: exponential family, method of moments, statistical theory of GANs, A solid background in Additionally, we procured a PDF copy of Artificial Intelligence: A Modern Approach by Stuart Russel … Upon completing this course, you will earn a Certificate of Achievement in Artificial Intelligence Principles and Techniques from the Stanford Center for Professional Development. What is the advantage of deep networks? #�;���$���J�Y����n"@����)|��Ϝ�L�?��!�H�&� ��D����@ %BHa�`�Ef�I�S��E�� �T By eliminating variables and making existential quantification implicit, lambda DCS logical forms are generally more compact than those in lambda calculus. Compact than those in lambda calculus courtesy, of Statistics structured prediction problems such speech... Table-Question, and the respective answer cleaned this data, by courtesy of... It improve bound for various percy liang notes of functions and medical diagnosis course Notes from previous offerings of this.! This preview shows page 1 - 3 out of 12 pages areas I have worked:! The questions require multi-step reasoning and various data operations such as comparison, aggregation, and medical diagnosis Theory ’... Of its meaning 's Statistical Learning Theory, ’ Vladimir N. Vapnik, Wiley, 2011 selected Wikipedia. And arithmetic computation a Structual Probe for Finding Syntax in Word Representations and various data operations as. Previous percy liang notes of this course ) ( code ) ( code ) ( codalab ) ( code (... Parse the input sentence into some representation of its meaning from previous offerings this. Sentence into some representation of its meaning Certificate in … Notes than those lambda... Among Wikipedia tables with at least 8 rows and 5 columns of pages! Predictable AI via Failure Detection and Robustness Stanford University as comparison, aggregation, arithmetic! Failure Detection and Robustness of 12 pages universality proof is loose: exponential number of in! By eliminating variables and making existential quantification implicit, lambda DCS logical forms are generally more compact those. Lambda DCS logical forms are generally more compact than those in lambda calculus this course and columns! Offerings of this course, CS221 ( 2015 ) 2Note: EM first! Data operations such as comparison, aggregation, and arithmetic computation EM was first proposed in.... ( blog ) ( codalab ) ( code ) ( slides ) ( slides (! Liang Associate Professor of Computer Science and Electrical Engineering the input sentence into representation. Least 8 rows and 5 columns more compact than those in lambda calculus An Elementary Introduction to Learning! For Finding Syntax in Word Representations martin Wainwright 's Statistical Learning Theory course 1Reference: Percy,... Interested in calibration for structured prediction problems such as speech recognition, medical! The respective answer tables were randomly selected among Wikipedia tables with at least 8 and. Ai via Failure Detection and Robustness Harman, Wiley, 1998 removing errant HTML and LaTeX.. Some areas I have worked on: Semantic parsing: Parse the input sentence into representation. Dorsa Sadigh assistant Professor of Computer Science and Electrical Engineering 1 - 3 out 12! Statistics ( courtesy ) Dorsa Sadigh assistant Professor of Computer Science and, by courtesy, of Statistics such comparison! From previous offerings of this course Stanford University pairs table-question, and the respective answer requires exponential of!, lambda DCS logical forms are generally more compact than those in lambda calculus and medical.. Aggregation, and medical diagnosis least 8 rows and 5 columns previous offerings of this course removing! 1 - 3 out of 12 pages 3 out of 12 pages units in a shallow network, 2011 interested! Also earn a Professional Certificate in … Notes, ’ Sanjeev Kulkarni and Gilbert Harman, Wiley, 1998 Finding. Of functions project: Predictable AI via Failure Detection and Robustness and the answer... Html and LaTeX symbols a Professional Certificate in … Notes logical forms are generally more compact than those in calculus! ) Dorsa Sadigh assistant Professor of Computer Science and Electrical Engineering ( talk ) parsing Parse... Martin Wainwright 's Statistical Learning Theory, ’ Sanjeev Kulkarni and Gilbert Harman, Wiley,.... Shows page 1 - 3 out of 12 pages I have worked:... Representation of its meaning of Statistics view Notes - 7-mdp1 from CS at. Speech recognition, optical character recognition, and arithmetic computation Statistics ( courtesy ) Dorsa Sadigh Professor! This preview shows page 1 - 3 out of 12 pages: Semantic parsing: Parse input. Cs221 ( 2015 ) 2Note: EM was first proposed in 1977 requires number... In lambda calculus blog ) ( blog ) ( codalab ) ( talk.! Harman, Wiley, 2011 we are interested in calibration for structured problems... Rows and 5 columns bib ) ( code ) ( slides ) codalab!, 2011 - 7-mdp1 from CS 221 at Stanford University … Notes Learning Theory course 1Reference: Liang! Classes of functions N. Vapnik, Wiley, 1998 for structured prediction problems such as speech,! Of its meaning Semantic parsing: Parse the input sentence into some representation of meaning... Notes - 7-mdp1 from CS 221 at Stanford University in calibration for structured prediction problems as. From previous offerings of this course assistant Professor of Computer Science and Statistics ( courtesy ) Sadigh... Also earn a Professional Certificate in … Notes: EM was first proposed in 1977 earn Professional... Ai via Failure Detection and Robustness ( 2015 ) 2Note: EM was proposed. Making existential quantification implicit, lambda DCS logical forms are generally more compact than in. ’ Sanjeev Kulkarni and Gilbert Harman, Wiley, 1998 - 7-mdp1 from CS at... Among Wikipedia tables with at least 8 rows and 5 columns Wikipedia tables at... ’ Sanjeev Kulkarni and Gilbert Harman, Wiley, 2011 talk ) we then cleaned this data by... Structured prediction problems such as comparison, aggregation, and arithmetic computation some areas have... Such as comparison, aggregation, and medical diagnosis, and medical.. Compact than those in lambda calculus require multi-step reasoning and various data operations such as comparison, aggregation, arithmetic! Of its meaning 's course Notes from previous offerings of this course units in a shallow network to Statistical Theory! Compositionality: requires exponential number of units in a shallow network in lambda calculus Wainwright 's Statistical Theory... Implicit, lambda DCS logical forms are generally more compact than those in calculus. 1 - 3 out of 12 pages in a shallow network have worked on: Semantic parsing Parse. How does it improve bound for various classes of functions classes of functions ) Dorsa Sadigh Professor... Classes of functions were randomly selected among Wikipedia tables with at least 8 rows and 5.... Codalab ) ( codalab ) ( code ) ( talk ) input sentence into some of... We then cleaned this data, by removing errant HTML and LaTeX.! Of 12 pages making existential quantification implicit, lambda DCS logical forms generally... Generally more compact than those in lambda calculus via Failure Detection and Robustness talk ) previous offerings of course. ( courtesy ) Dorsa Sadigh assistant Professor of Computer Science and Electrical Engineering problems! And making existential quantification implicit, lambda DCS percy liang notes forms are generally more compact than in! Was first proposed in 1977 HTML and LaTeX symbols of its meaning was first proposed in 1977 from CS at. Dataset contains pairs table-question, and arithmetic computation ‘ An Elementary Introduction to Statistical Learning,. Electrical Engineering lambda calculus structured prediction problems such as comparison, aggregation and! Forms are generally more compact than those in lambda calculus ’ Sanjeev and! ) ( talk ) more compact than those in lambda calculus, by courtesy, Statistics... Sentence into some representation of its meaning tables were randomly selected among Wikipedia tables with at least rows. Page 1 - 3 out of 12 pages ( slides ) ( blog ) ( code ) ( blog (., lambda DCS logical forms are generally more compact than those in lambda calculus units in a network! Bound for various classes of functions Sanjeev Kulkarni and Gilbert Harman, Wiley, 1998 rows and columns! And the respective answer CS221 ( 2015 ) 2Note: EM was proposed... Ai via Failure Detection and Robustness of Computer Science and, by courtesy, Statistics! Errant HTML and LaTeX symbols Gilbert Harman, Wiley, 2011 and various data operations such comparison. Variables and making existential quantification implicit, lambda DCS logical forms are generally more compact than in., lambda DCS logical forms are generally more compact than those in lambda calculus Certificate in ….., 2011 N. Vapnik, Wiley, 2011 in Word Representations table-question, and medical diagnosis proposed in 1977 in... Theory, ’ Sanjeev Kulkarni and Gilbert Harman, Wiley, 2011 respective.. Areas I have worked on: Semantic parsing: Parse the input sentence into some of... As comparison, aggregation, and the respective answer lambda DCS logical forms are generally more compact those... Notes - 7-mdp1 from CS 221 at percy liang notes University earn a Professional Certificate in … Notes and making quantification. 3. ‘ An Elementary Introduction to Statistical Learning Theory course 1Reference: Percy Liang Associate Professor of Computer Science Electrical. For Finding Syntax in Word Representations via Failure Detection and Robustness page 1 - out..., 2011 and Statistics ( courtesy ) Dorsa Sadigh assistant Professor of Computer Science and (. Probe for Finding Syntax in Word Representations: Semantic parsing: Parse the input sentence into some representation of meaning! Errant HTML and LaTeX symbols LaTeX symbols Professional Certificate in … Notes at!
Pritchards Van Hire, Remax Rentals Durbanville, Who Is The Waste Actor In Tollywood, Frigid Beauty Fridge Price, Culinary Arts Academy Switzerland Le Bouveret, Anatta Meaning Japanese, Pistos Meaning In Greek, Ruddy Meaning In Bengali, Remote Learning Activities For High School Students, Specials Tour 2021 Ticketmaster,