Courses in Master of Artificial Intelligence

Courses description

ICS 501: Foundations of Artificial Intelligence (3-0-3)

Fundamental concepts and techniques of intelligent systems. Principles and methods for heuristic search, knowledge representation, problem solving, planning and reasoning with uncertainty, game and adversarial search and their application to building intelligent systems in a variety of domains. Basics of machine learning, visual perception and natural language processing. Introduction to AI programming.

Note: Cannot be taken for credit with ICS-381.

ICS 502: Machine Learning (3-0-3)

Introduction to machine learning; supervised learning (linear regression, logistic regression, classification, support vector machines, kernel methods, decision tree, Bayesian methods, ensemble learning, neural networks); unsupervised learning (clustering, EM, mixture models, kernel methods, dimensionality reduction); learning theory (bias/variance tradeoffs); and reinforcement learning and adaptive control.

Note: Cannot be taken for credit with ICS-485.

ICS 503: Evolutionary Computation and Global Optimization (3-0-3)

An introduction to a wide variety of robust optimization algorithms based on the theme of nature inspired optimization techniques. Computational implementation single-state methods such as Simulated Annealing and Tabu Search ; and population-based methods such as Genetic Algorithms, Particle Swarm, and Ant Colony. Theory including representations, landscapes, epistasis, code bloat, diversity, and problem structure is discussed. Applications to optimization, machine learning, software development, and others.

ICS 504: Deep Learning (3-0-3)

Deep Learning models and their applications in real world. Foundations of deep learning networks training and optimization. Deep learning models for spatial and temporal data processing. Analysis of prominent deep learning models such as Convolutional Neural Networks (CNNs), Recurrent and Recursive Networks, Long-Short Term Memory (LSTM), Residuals Networks, and Generative Adversarial Networks (GANs). One-Shot Learning and Deep Reinforcement Learning.

Note: Cannot be taken for credit with ICS-471.

ICS 505: Computer Vision (3-0-3)

Taxonomy of computer vision tasks, Applications of computer vision, Image representation in the spatial and frequency domains, Image formation, Image filtering, Feature detection and matching, Image segmentation, Image classification, Object detection, Image alignment and stitching, Motion estimation and tracking, Depth estimation, Deep learning for computer vision.

Note: Cannot be taken for credit with ICS-483.

ICS 506: Natural Language Processing (3-0-3)

Natural language processing (NLP) fundamentals, Language modeling, Vector space semantics and Embeddings, Sequence labelling, Syntactic parsing, semantic analysis, Information Extraction, Machine translation, Discourse Coherence, Question Answering, Dialogue Systems and Chatbots, and Natural language summarization.

Note: Cannot be taken for credit with ICS-472.

ICS 619: Project (6-0-6)

A graduate student will arrange with a faculty member to conduct an industrial research project related to the Artificial Intelligence and Machine Learning field of the study. Subsequently the students shall acquire skills and gain experiences in developing and running actual industry-based project. This project culminates in the writing of a technical report, and an oral technical presentation in front of a board of professors and industry experts.

MATH 503: Mathematics for Data Science (3-0-3)

Selected topics from linear algebra, multivariate calculus, and optimization for Data Science with an emphasis on the implementation using numerical and symbolic software, toolboxes, and libraries for data science like NumPy, SciPy, Pandas, SymPy. Topics include data transformation using linear algebra, vector spaces, linear transformations, matrix representations, matrix decompositions (eigenvectors, LU, QR, SVD, Cholesky); multivariate calculus for continuous, convex, and non-convex optimization methods; time series construction and visualization, Fourier transformations for time series conversion.

MATH 506: Fundamentals of Data Science (3-0-3)

All aspects of the data science pipeline using the software, toolboxes, and libraries like NumPy, SciPy, Pandas, Matplotlib, Seaborn: data acquisition, cleaning, handling missing data, EDA, visualization, feature engineering, modeling, model evaluation, bias-variance tradeoff, sampling, training, testing, experimenting with a classical model; ethics in data science.