Methods III: Statistical Machine Learning

Catalog description

This course introduces modern statistical learning tools with a specific focus on those developed for big data. It covers three interconnected components: statistical machine learning methods, the underlying algorithms, and computational tools. The primary emphasis is on key techniques for comprehensive data analysis, including managing large datasets, exploring patterns, framing statistical problems, constructing efficient computational algorithms, and generating reports. Topics encompass data management, feature engineering, clustering, convex optimization algorithms, tree/ensemble methods, and predictive modeling. Upon completion, participants should be able to (a) manage and explore large data sources using techniques such as sampling, parallel processing, or other tools, (b) apply, evaluate, and interpret results from statistical learning tools, including clustering, decision trees, and neural networks, (c) understand the underlying assumptions and mathematical foundations of these methods, discerning how associated parameters influence outcomes, (d) compare the computational complexity of different statistical