About

Table of contents

  1. Course structure
  2. Grading policy
  3. Late submissions
  4. Exit passes
  5. Lecture scribing
  6. Academic integrity
  7. Students with disability
  8. Textbooks

Course structure

  1. Overview & Basics
    1. Introduction to clinical care and clinical data
      • Goals of clinical care
      • History and unique challenges
      • Attributes of clinical data
    2. Math and python refresher
      • Probability, linear algebra, statistics
      • NumPy, Pandas, Scikit-Learn, PyTorch
  2. Classification Problems
    1. Classification methods in healthcare
      • Diagnosis, prognosis, stratification, segmentation
    2. Different model choices and interpretation
      • Logistic regression, SVM, random forests, xgboost
      • Hyperparameters
      • Evaluation
  3. Unsupervised Problems
    1. Unsupervised methods for healthcare
      • Subtype discovery, pattern discovery
    2. Dimensionality reduction
      • PCA, ICA, tensor decomposition, t-SNE
      • How and when to use them?
    3. Clustering methods
      • k-means, mixture models, hierarchical clustering
      • Understanding, and evaluating clustering results
  4. Deep Learning
    1. Convolutional neural networks
      • Medical imaging
      • Physiological signals
    2. Graphs and graph neural networks
      • Healthcare knowledge graphs
    3. Sequence models in healthcare
      • Hidden Markov models
      • Recurrent neural networks, LSTM, Attention
    4. Unsupervised deep learning for healthcare
      • Autoencoders
      • Self-supervised learning
  5. Advanced Topics
    1. Generative models in healthcare
    2. Clinical foundation models
    3. Federated learning
    4. Trustworthy AI
      • Robustness
      • Fairness
      • Safety
      • Ethics

Grading policy

CategoryCSCI 5980 sectionCSCI 8980 section
Machine Problems60% ( 3 * 20%)45% ( 3 * 15%)
Final Project40%40%
Mini Project-15%

Late submissions

No late submissions are allowed. Extenuating circumstances require prior approval.

Exit passes

(To be updated)

Lecture scribing

(To be updated)

Academic integrity

As this is an advanced graduate-level course, students are expected to uphold the highest standards of academic integrity. Discussion of ideas is allowed for machine problems and lecture scribing assignments, but the submitted work must be your own. For each problem, please do cite any resources that you have used to solve it. Use of generative AI content in assignments is not allowed. Cheating in this course will result in a failing grade (F), and the University’s policies on academic dishonesty will be strictly enforced.

Students with disability

Students requiring accommodations are encouraged to discuss their needs with the instructor at the beginning of the semester or as soon as possible. The course staff is committed to ensuring an inclusive and accessible learning environment for all students.

Textbooks

There are no recommended textbooks for the class. However few topics will be borrowed from

  • The Elements of Statistical Learning: Data Mining, Inference, and Prediction, Second Edition. Trevor Hastie, Robert Tibshirani, and Jerome Friedman.
  • Deep learning. Ian Goodfellow, Yoshua Bengio and Aaron Courville.
  • Deep Learning: Foundations and Concepts. Christopher Bishop and Hugh Bishop.