Computer Science: Machine Learning

Introduction to machine learning algorithms, theory, and implementation, including supervised and unsupervised learning; topics typically include linear and logistic regression, learning theory, support vector machines, decision trees, backpropagation artificial neural networks, and an introduction to deep learning. Includes a team-based project. Credit not awarded for both CS 3540 and CS 5540. Prerequisites: STAT 2510 or STAT 5510; MATH 2522 or MATH 2544.

Prereqs enforced by the system: STAT 2510 or 5510, and MATH 2522 or 2544; Co-located with CS 5540 , CSYS 5990, Total combined enrollment: 50; Open to Degree and PACE students;

This course offers an introduction to machine learning and statistical pattern recognition algorithms. It includes a wide array of topics, including supervised learning (covering linear regression, logistic regression, neural networks, support vector machines, decision trees, ensemble models, and random forests), unsupervised learning (clustering, dimensionality reduction, principal component analysis, and kernel methods), and deep learning (featuring convolutional neural networks, basic recurrent neural networks, and LSTMs). We will also introduce data-driven modeling techniques which are useful in the engineering and physical science domains (e.g., physics informed neural networks, neural ODEs, Bayesian neural networks, etc). You will also gain valuable insights and guidance on effectively implementing machine learning algorithms.

Weekly assignments, Term Project and two Mid-term exams.

• Assigned homework (30%) • Term Project (30%) • Mid term exams (40%). Two exams will be given during the semester, no final exam.