CSC/EE 8001: End-to-End ML Systems
Machine learning systems are both complex and unique. Complex because they must balance both performance and accuracy. Unique because they’re data dependent, with data varying wildly from one use case to the next. In this course we will learn why and how to build systems that learn from data. We will learn that it requires more than understanding algorithms—it demands expertise spanning data pipelines, model development, optimization for deployment constraints, and operational practices. We will acquaint ourselves with both compute and data issues arising in managing a large ML pipeline. The course will be project-based.