## Courses

Notes: Syllabi, Lecture slides, Homeworks, and Projects will be distributed through Tulane Canvas website.

### CSE 5526: Introduction to Neural Networks

- Course description: Survey of fundamental methods and techniques of neural networks; single- and multi-layer perceptrons; radial basis function networks; support vector machines; recurrent networks; supervised and unsupervised learning
- Course level: Undergraduates (seniors), and graduates (masters and doctoral)
- Textbook: (Recommended) Neural networks and learning machines Simon Haykin
- Topics covered:
- Multi-layer perceptrons
- Radial-basis function networks
- Deep neural networks
- Recurrent and dynamic networks
- Various applications

### CSE 5523: Machine Learning and Statistical Pattern Recognition

- Course description: Introduction to basic concepts of machine learning and statistical pattern recognition; techniques for classification, clustering and data representation and their theoretical analysis.
- Course level: Graduate/Undergrad
- Prerequisites: (CSE 3521 or CSE 5521 or CSE 5243) and (CSE 5522 or Stat 3460 or Stat 3470)
- Topics covered:
- Probability basics
- Introduction to machine learning: What is machine learning? Linear regression, Minimizing training error, linear Gaussian model, Mean-square error, Bias-Variance decomposition, Probability of error, Bayes decision rule, Linear Discrimant Analysis, Chernoff bound
- Empirical risk minimization: Expected and empirical risks, Surrogate losses, PAC-learnability, Sample complexity, Reliazable vs unrealizable, Markov inequality, Hoeffding's inequality, VC-dimenion, Generalization bound
- Convex Optimization: Hessian, PSD, Local and global minima, Optimality condition, Covex sets and functions, Subgradient, Convex optimization problem, Gradient descent, Newton's method, Convergence, Batch vs stochastic, Constrained optimization, Lagrange multipliers, KKT conditions, Dual functions and problems, Duality
- SVM and Kernel methods: Maximum margin, SVM in primal and dual forms, KKT condition, Support vectors, Soft-margin SVMs, Hinge loss, Polynomial feature map, Kernels, PSD, Gaussian kernels, Closedness, RKHS, Representer theorem, Kernel trick, Kernel-SVM/RR/LDA, Nystrom approximation, Random Fourier features, Other kernels
- Dimensionality reduction: Eigen-decomposition, PCA, Kernel PCA, MDS, Isomap, LLE, Laplacian Eigenmap, K-means, Spectral clustering
- Other topics

- References: There is no required textbook. Homeworks and exams will be based on provided lecture notes, which are partially based on the following references.
- Stuart Russell and Peter Norvig. Artificial Intelligence: A Modern Approach (3rd Edition), Prentice Hall, 2009.
- Christopher M. Bishop. Pattern Recognition and Machine Learning, Springer, 2006.
- Shai Shalev-Shwartz and Shai Ben-David. Understanding Machine Learning: From Theory to Algorithms, Cambridge University Press
- Mehryar Mohri, Afshin Rostamizadeh, and Ameet Talwalkar, Foundations of Machine Learning, MIT Press, 2012.
- Bernhard Schölkopf and Alexander J. Smola, Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond, MIT Press, 2001
- Daphne Koller and Nir Friedman. Probabilistic Graphical Models: Principles and Techniques, MIT Press, 2009.

### CSE 5522: Survey of Artificial Intelligence II: Advanced Techniques

- Course description: Survey of advanced concepts, techniques, and applications of artificial intelligence, including knowledge representation, learning, natural language understanding, and vision.
- Course level: Undergraduates/Graduates
- Prerequisites: CSE 3521 or CSE 5521 or CSE 630 or grad standing
- Topics covered:
- Probability review
- Bayesian networks
- Markov networks
- Statistical estimation
- Mixture models
- EM algorithm
- Hidden Markov Models
- Intro to Machine Learning
- Support Vector Machines
- Neural Networks
- Optimization
- Applications

- References: There is no required textbook. Homeworks and exams will be based on provided lecture notes, which are partially based on the following references.
- Stuart Russell and Peter Norvig. Artificial Intelligence: A Modern Approach (3rd Edition), Prentice Hall, 2009.
- Christopher M. Bishop. Pattern Recognition and Machine Learning, Springer, 2006.
- Daphne Koller and Nir Friedman. Probabilistic Graphical Models: Principles and Techniques, MIT Press, 2009

### CSE 3521: Survey of Artificial Intelligence I: Basic Techniques

- Course Description: A survey of the basic concepts and techniques of problem solving and knowledge representations in AI.
- Course level: Undergraduates (seniors), and graduates (masters with a non-AI focus)
- Textbook: Artificial Intelligence: A Modern Approach by Russell and Norvig, third edition, Prentice Hall, 2010. More information on the book is found at http://aima.cs.berkeley.edu/.
- Topics covered:
- Introduction and course overview (Ch 1)
- Intelligent agents (Ch 2)
- Problem-solving agents (Ch 3.1-3.3)
- Uninformed searching (Ch 3.1-3.4)
- Informed searching (Ch 3.4-3.6)
- Local search and optimization (Ch 4.1 – 4.2)
- Searching with nondeterministic actions/partial observations (Ch 4.3-4.4)
- Games (Ch 5)
- Propositional logic (Ch 7-8)
- First-order logic (Ch 9),
- Probability and Bayes’ rule (Ch 13-14)
- Bayesian network (Ch 15)
- Utility theory and MDP (Ch 16, 17)
- Machine learning 1 (Ch 18)
- Vision and robotics (Ch 24, 25)
- Natural language processing (Ch 22-23)