THEORETICAL FOUNDATIONS OF LARGE-SCALE MACHINE LEARNING
Mathematical foundations of large-scale machine learning and optimization. Focus on recent texts in machine learning, optimization, and randomized algorithms, focused on tradeoffs that are driving algorithmic design in this new discipline. These trade-offs revolve around speed of convergence, statistical accuracy, robustness, scalability, algorithmic complexity, and implementation.
Prerequisites & course information
Requisites: COMP SCI/E C E 761 Course Designation: Grad 50% - Counts toward 50% graduate coursework requirement Repeatable for Credit: No Last Taught: Spring 2024 Learning Outcomes: 1. Prove convergence rates for stochastic optimization algorithms Audience: Graduate 2. Describe systems tradeoffs that drive algorithmic design in large-scale machine learning Audience: Graduate 3. Summarize recent literature in large-scale optimization and machine learning Audience: Graduate 4. Design, tune, or tailor a given machine learning algorithm for a new system Audience: Graduate 5. Evaluate statements about parallelizability, generalization, convergence of optimization algorithms Audience: Graduate Learning Outcomes: 1. Prove convergence rates for stochastic optimization algorithms Audience: Graduate 2. Describe systems tradeoffs that drive algorithmic design in large-scale machine learning Audience: Graduate 3. Summarize recent literature in large-scale optimization and machine learning Audience: Graduate 4. Design, tune, or tailor a given machine learning algorithm for a new system Audience: Graduate 5. Evaluate statements about parallelizability, generalization, convergence of optimization algorithms Audience: Graduate
Sources
- Course catalogue ↗2026-2027