TTIC Courses. Held in cooperation with the University of Chicago Department of Computer Science. Autumn 2. 01. 7TTIC 3. Introduction to Statistical Machine Learning 1. Greg Shakhnarovich TTIC Room 5. B TR 2 0. 0 3 2. Midterm Exam will be held not at TTIC, but at SS 1. See Instructor or TAs for details. Final Exam Room TBD Tuesday, December 5, 1 3. PMTTIC 3. 11. 50 Mathematical Toolkit CMSC 3. Mesrob Ohannessian TTIC Room 5. T,R 9 3. 0 1. 0 5. Final Exam Room 5. Thursday, December 7, 1. AM 1. 2 3. 0 PMTTIC 3. Information and Coding Theory CMSC 3. Madhur Tulsiani TTIC Room 5. M,W 1 3. 0 2 5. Final Exam Rooms 5. Wednesday, December 6, 1 3. PMTTIC 3. 12. 40 Self driving Vehicles Models and Algorithms for Autonomy 1. Matthew Walter TTIC Room 5. M,W 9am 1. 1am. Course Website. TTIC 3. 10. 00 Research at TTICFridays, Room 5. TTIC 5. 50. 00 Independent Research 1. TTIC 5. 60. 00 Independent Reading 1. TTIC 5. 70. 00 Computer Science Internship. Complete Course List. Web form. 10. 0 units. TBDWeekly lectures and discussions by TTIC researchers introducing their research and research problems. Introduction To Algorithms Third Edition PdfProvides a broad view of research carried out at TTIC. Course is passfail credit. Satisfies one quarter of credit of the three required to fulfill the Research at TTIC Series Requirement. See Academic Program Guide for details1. Chuzhoy, Julia, and Makarychev, Yury. This is a graduate level course on algorithms with the emphasis on central combinatorial optimization problems and advanced methods for algorithm design and analysis. Topics covered include greedy algorithms, dynamic programming, randomized algorithms and the probabilistic method, combinatorial optimization and approximation algorithms, linear programming, and online algorithms. The course textbook is Algorithm Design by Kleinberg and Tardos. Lecture Plan. Greedy algorithms 1 weekDynamic programming 1 weekOnline algorithms 1 week. Max flow, min cut, bipartite matching and their applications 3 weeksLinear programming, LP duality 1 weekNP hardness 1 weekApproximation algorithms 1 weekRandomized algorithms 1 weekAssumes familiarity with proofs and an the asymptotic notation. Some basic knowledge of the notion of NP hardness is also required. Expected outcomes Ability to design and rigorously analyze algorithms using paradigms such as greedy or dynamic programming. Understand the use of linear programming in optimization. Be able to formulate problems as linear programs. Understand linear programming duality and applications to problems such as max flowmin cut. Be able to write duals for linear programs. Introduction to Algorithms is a book by Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest, and Clifford Stein. The first edition of the book was widely used as. Introduction to genetic algorithms, tutorial with interactive java applets, Introduction. Depth The computations involved in producing an output from an input can be represented by a flow graph a flow graph is a graph representing a computation, in. An overview of a network as a collection of connected elements. Different types of networks are illustrated as well as a way to represent them mathematically. This course provides an introduction to mathematical modeling of computational problems. It covers the common algorithms, algorithmic paradigms, and data structures. 3D Puzzle Dxf. Introduction to genetic algorithms, tutorial with interactive java applets, Main page. Algorithms have uses in computer programming and beyond Find out what algorithms are and how to write and use them with Flocabularys educational hiphop song and. Prerequisites Assumes familiarity with proofs and an the asymptotic notation. Some basic knowledge of the notion of NP hardness is also required. Shakhnarovich, Greg. A systematic introduction to machine learning, covering theoretical as well as practical aspects of the use of statistical methods. Topics include linear models for classification and regression, support vector machines, regularization and model selection, and introduction to structured prediction and deep learning. Application examples are taken from areas like information retrieval, natural language processing, computer vision and others. Prerequisites Probability, Linear Algebra, Undergraduate Algorithms. Lecture Plan. I will focus on supervised learning and only talk about unsupervised settings when necessary e. So, no clustering. There will be twenty 1. ML, motivation etc. MLMAP, overfitting, biasvariance 1model complexity, sparsity L1L2 in regression stepwise methods for L0 sparsity 1classification Fishers LDA, logistic regression and softmax 1ensemble methods, boosting 1generative models, Naive Bayes, multivariate Gaussians 1mixture models EM 2SVM and kernels 2nonparametric methods nearest neighbors, density esimation 1multilayer neural networks and deep learning 1information theory and learning information criteria, MDL and their connections to regularization 1experiment design and evaluation in ML 1advanced topics TBD 1wrap up and review 1Prerequisites knowledge of basic linear algebra, probability and calculus. Expected outcomes Understand the notion of fitting a model to data and concepts such as model complexity, overfitting and generalization, and bias variance tradeoff in estimation. Learn and be able to apply some of the fundamental learning methods, such as logistic regression, support vector machines, boosting, decision trees, neural networks. Learn the basics of optimization techniques such as gradient descent and the general EM algorithm. Familiarity with multivariate Gaussians and mixtures of Gaussians. Understand fundamental concepts in information theory entropy, KL divergence and their relationship to machine learning. Mc. Allester, David. This course covers the foundations of mathematics from a classical nonconstructive type theoretic perspective, the general notion of a mathematical structure, the general notion of isomorphism, and the role of axiomatizations and constructions in mathematical definitions. The definition of the real numbers used as a fundamental example. The course also covers the notion of definability in well typed formalisms. Brian Eno Cards Oblique Strategies Pdf Printer. A primary example is the non definability of a linear bijection between a vector space and its dual. Ontologies types relevant to machine learning are emphasized such as the type theory of PCA, CCA and Banach spaces norms and dual norms. Lecture Plan. There are again two lectures per week. Introduction and course outline. Part I Logic with Abstraction Barriers. Sequent inference rules and proofs. Types and variable declarations. Free and bound variables. Structures and Isomorphism. Isomorphism as equivalence under an abstract interface. Part II Case Studies in Abstraction. The natural numbers, integers, rationals and reals. Install Curl Windows Server 2003 more. Vector spaces. A formal treatment of the non existence of a canonical basis coordinate system, canonical inner product, or canonical isomorphism with the dual space. Coordinate free treatment of matrix algebra. Equivalences between matrix operator types. The fact that least squares regression does not require an ambient inner product but regularization does. Gradient descent. Gradient descent requires an ambient inner product. Newtons method does not. The covariance matrix of a probability distribution on a vector space. The fact that the multivariate central limit theorem does not require an ambient inner product. Canonical correlation analysis also does not require an ambient inner product. PCA requires and ambient inner product. Norms and Banach spaces. Dual norms. Measure spaces. Hilbert space. Differentiable manifolds. Information geometry. Jefferys prior. Natural Gradient Descent. Expected outcomes Be able to write rigorous proofs and reason formally about various mathematical notions. Understand basics of linear algebra, eigenvalues and eigenvectors from the viewpoint of operators. Understand various algorithms such as SVMs, PCA, CCA and gradient descent from the operator viewpoint. Mc. Allester, David. Introduction to deep learning for computer vision. Although deep learning based computer vision systems are evolving rapidly, this course attempts to teach material that will remain relevant and useful as the field changes. The course begins with general deep learning methods relevant to many applications and gradually focuses to a greater extent on computer vision.