Programming for AI/ML

Catalog description

The objective of this course is to familiarize students with the frameworks and ecosystems for building end-to-end Artificial Intelligence (AI) and Machine Learning (ML) Pipelines. Students will be exposed to python-based data processing, data analytics and machine learning libraries such as pandas, sci-kit; and deep learning libraries such as Pytorch. The course will also introduce the students to the basics of accelerating AI/ML programs to using GPUs, distributed training and using KV Cache optimizations along with Agentic AI. Finally, the course will introduce python based GPU programming using Triton. The graduate version of this course will allow students to work on a project to implement and enhance an end to end AI/ML pipeline for a real world inspired use-case scenario. Offered as CSDS 350 and CSDS 450 . Prereq: ( CSDS 134 or CSDS 237 ) and CSDS 233 .