This class covers the following topics: Introduction to smart healthcare; health decision support system; wearable medical sensors and deep neural network based disease detection; continual learning based multi-headed neural networks for multi-disease detection; interpretability through differentiable logic networks; interpretability through conformal predictions; medical images and convolutional neural network based disease detection; natural language processing for healthcare; foundation models for healthcare; counterfactual reasoning based personalized medical decision-making.
All assignments are built on papers you will read for the class. Each of these coding tasks exemplifies the paper discussed this week, and reproduces a simplified version of the smart healthcare framework in this paper. The assignments introduce you to the latest advances in Machine Learning (broadly) and especially Deep Learning.
Useful reference to ML concepts: ML Glossary from Google
- Instructor: Niraj K. Jha
- TAs: Yuval Kansal
- Lectures: M/W 10:40am-12:00pm
- Office hours:
- Yuval Kansal: Tu: 3:30-4:30pm, Th: 11am-12pm (EQuad B321)
Each assignment is worth 20 points. You are HIGHLY encouraged to use Google Colab, especially if you are unable to run the code locally. Detailed instructions on using Colab can be found here:
- [One-minute Tutorial](https://drive.google.com/file/d/1ANaoLv6HWze6RLm1Fvsv8qpHpM4GZGQb/view?usp=drive_link)
BSD-3-Clause. Copyright (c) 2026, JHA-Lab. All rights reserved.
See License file for more details.