Probabilistic Machine Learning. 3 credits

Catalog description

Machine learning methods rely on probabilistic and statistical models to perform various tasks. The applications range from social media features to sentiment analysis and healthcare efficiency. This course will train the students in powerful probabilistic methods and computations that will enable them to perform large and complex data analysis projects. The course will cover materials on both supervised and unsupervised learning. The course will begin with a quick review of probability models and statistical inference involving frequentist and Bayesian decision theory. This would be followed by probabilistic models for machine learning, including hidden Markov models, probabilistic graphical models that include Bayes nets, and related concepts such as belief propagation on graphs (sum-product algorithm), Regression, regularization, EM algorithm, Kernels, and Uncertainty quantification. Offered by Statistics and cross-listed with ECE 657 . May not be repeated for credit. Offered by Statistics . May not be repeated for credit.