Model explanability - One of the biggest problems in machine learning (ML) is understanding what a model is actually doing and how it arrives at results. Students could explore and analyze ML models similar to
and interpret what features the model is looking for.
Creating an Alexa-type personal assistant / chatbot - in the process of designing and coding this service, students will learn natural language processing, which is an important branch of AI. To make sure their bot understands what a user wants, students will need to deeply learn grammar.