(3 units). Bayesian networks, factor graphs, Markov random fields, maximum a posteriori probability (MAP) and maximum likelihood (ML) principles, elimination algorithm, sum-product algorithm, decomposable and non-decomposable models, junction tree algorithm, completely observed models, iterative proportional fitting algorithm, expectation- maximization (EM) algorithm, iterative conditional modes algorithm, variational methods, applications. Courses CSI 5149 (COMP 5007), ELG 5131 (EAGJ 5131) and ELG 7177 (EACJ 5605) cannot be combined for units. This course is equivalent to COMP 5007 at Carleton University. Course Component: Lecture Permission of the Department is required.
(3 units). Bayesian networks, factor graphs, Markov random fields, maximum a posteriori probability (MAP) and maximum likelihood (ML) principles, elimination algorithm, sum-product algorithm, decomposable and non-decomposable models, junction tree algorithm, completely observed models, iterative proportional fitting algorithm, expectation- maximization (EM) algorithm, iterative conditional modes algorithm, variational methods, applications. Courses CSI 5149 (COMP 5007), ELG 5131 (EAGJ 5131) and ELG 7177 (EACJ 5605) cannot be combined for units. This course is equivalent to COMP 5007 at Carleton University. Course Component: Lecture Permission of the Department is required.
