Yale School of Public Health · Biostatistics

Introduction to
Health Data Science

Turn health questions into precise problems. Develop the tools to analyze data, evaluate evidence, and explain what you find.

Open Canvas Find your next class
  1. 01
    Data structures & algorithmsRepresent a problem. Design a solution.
  2. 02
    Machine learning & AIFit, compare, and evaluate predictions.
  3. 03
    Causal inferenceDefine effects. Examine assumptions.
Your course desk

Coming up

Next class

Find your next class and its reading in the lecture calendar.

Winslow Auditorium, LEPH 109, 60 College St. Bring your name card; every class has a short paper quiz.

Class-wide deadlines

Scribe deadlines are in the lecture calendar, and all course dates are under Key Dates. Where this page and Canvas disagree, Canvas has the final deadline.

01 / Templates & downloads 02 / Readings & scribe notes 03 / Get help in office hours
Prepare · attend · review

Lecture calendar

Lecture Calendar

26 lectures, Thu Sep 3 to Thu Dec 10, following the Yale academic calendar. Each row lists the class topic, its reading, its scribe notes, and what is due. Topics are confirmed after each class, and a topic marked Planned has not been taught yet. The planned topic sequence for the term is below.

Reading links open the free PDF at the first page of the section. Deadlines are set on Canvas; where this page and Canvas disagree, Canvas is right.

Lecture calendar. For each class: the date, topic, reading, scribe notes, and deadlines or announcements.
ClassTopicReadingScribe notesDue and announcements
Lecture 1 What is Data Science?
Lecture 2 Algorithmic Thinking, Recursion, and Bubble Sort
Lecture 3 Merge Sort: Halting, Correctness, and Comparison Counts
Lecture 4 Computational Complexity: O, Ω, Θ, and Merge Sort’s Recurrence
Lecture 5 Graphs and Breadth-First Search: Whom Should We Test Next? Posted after grading
  • Project teams due Fri Sep 18
Lecture 6 Analysis of Breadth-First Search: Termination, Correctness, and Running Time Posted after grading
  • Project proposal instructions posted on Canvas
Lecture 7 Dynamic Programming: Designing a Longest Common Subsequence Algorithm, With and Without Memoization Posted after grading
Lecture 8 Analysis of the Longest Common Subsequence Algorithm Planned
Lecture 9 Topic posted after class
  • Proposal presentations in class: one speaker per team, five minutes
Lecture 10 Topic posted after class
Lecture 11 Topic posted after class
Lecture 12 Topic posted after class
Lecture 13 Topic posted after class
  • HW 2 due Sat Oct 17 on Canvas
Lecture 14 Topic posted after class
No classOctober recess
Lecture 15 Topic posted after class
Lecture 16 Topic posted after class
Lecture 17 Topic posted after class
Lecture 18 Topic posted after class
Lecture 19 Topic posted after class
Lecture 20 Topic posted after class
Lecture 21 Topic posted after class
Lecture 22 Topic posted after class
No classNovember recess
No classNovember recess
Lecture 23 Topic posted after class
Lecture 24 Topic posted after class
Lecture 25 Topic posted after class
Lecture 26 Last class; topic posted after class

Student scribe notes appear next to their lecture once they have been graded. Scribe assignments are on Canvas; the instructor examples and the template are under Course Files. Calendar last updated Sun Sep 27, 2026.

Three modulesThe course roadmap

Topics in the order they will be taught. Dates are set as the course goes; the calendar records what was actually covered. Readings are free online; see Materials.

M1: Data Structures & Algorithms
M2: Machine Learning & AI
M3: Causal Inference
n learning objective
  1. Module 1, data structures and algorithms: Course Introduction
    What data science is; problem formulation; translating questions to estimands
    Turing (1950) · Lec 1 note
    Learning objectives 12
  2. Module 1, data structures and algorithms: Algorithms, Sorting & Complexity
    Algorithmic thinking; recursion and induction; bubble and merge sort; O, Ω, Θ
    CLRS Ch. 2–4 · Lec 2 note · Lec 3 note
    Learning objectives 23
  3. Module 1, data structures and algorithms: Trees & Graph Traversal
    Tree structures; BFS; DFS; applications in healthcare networks
    CLRS Ch. 10, 22
    Learning objectives 23
  4. Module 1, data structures and algorithms: Dynamic Programming
    Optimization principles; memoization; tabulation; sequence alignment
    CLRS Ch. 15
    Learning objectives 23
  5. Module 2, machine learning and AI: Statistical Learning
    Maximum likelihood estimation; uncertainty quantification
    ISLR Ch. 1–2; IAML Ch. 1.1
    Learning objectives 14
  6. Module 2, machine learning and AI: Linear Methods
    Linear & logistic regression; regularization (Ridge, LASSO); evaluation metrics
    ISLR Ch. 3, 4, 6; IAML Ch. 4
    Learning objectives 134
  7. Module 2, machine learning and AI: Nonparametric Methods
    Decision trees; random forests; boosting & bagging
    ISLR Ch. 8; IAML Ch. 2–3
    Learning objectives 346
  8. Module 2, machine learning and AI: Neural Networks
    Perceptrons; multi-layer networks; backpropagation; deep learning intro
    ISLR Ch. 10; IAML Ch. 12.1
    Learning objectives 34
  9. Module 2, machine learning and AI: Clustering
    K-means; EM algorithm; hierarchical clustering
    ISLR Ch. 12.4; IAML Ch. 10
    Learning objectives 346
  10. Module 3, causal inference: Potential Outcomes & DAGs
    Rubin Causal Model; causal graphs; confounding; identification strategies
    Mixtape Ch. 3–4; FCCI Ch. 1–2
    Learning objectives 124
  11. Module 3, causal inference: Matching & Weighting
    Matching and weighting estimators; difference-in-differences
    Mixtape Ch. 5; FCCI Ch. 10–11
    Learning objectives 346
  12. Module 3, causal inference: Project Presentations
    Student poster session; data equity discussion
    Data Equity in Data Science
    Learning objectives 56
Files · books · software

Your course toolkit

Course Files

Homework

Course project

Scribe notes

On Canvas

Course flyer: web · PDF. Course policies in one place: Policies.

No single required textbook; every text below is free online. The lecture calendar lists the reading for each class. CLRS section numbers refer to the free 2nd edition linked here; other editions number some chapters differently, so look a section up by its title.

Introduction to Algorithms
Cormen, Leiserson, Rivest, Stein
Intuition for the Algorithms of ML
Cynthia Rudin · course page
An Introduction to Statistical Learning
James, Witten, Hastie, Tibshirani, Taylor · Python labs
A First Course in Causal Inference
Peng Ding · arXiv 2305.18793
Causal Inference: The Mixtape
Scott Cunningham · read online
Algorithms
Jeff Erickson (“Erickson” in the calendar) · free chapter PDFs
Mathematics for Computer Science
Lehman, Leighton, Meyer (2018 revision) · free PDF
Python Data Science Handbook
Jake VanderPlas · notebooks

Software Stack

Python 3.x NumPy Pandas PyTorch Scikit-learn Matplotlib Seaborn Jupyter Overleaf (LaTeX)

Install and compile instructions are in the Homework 1 handout (Getting started: Python and LaTeX). Overleaf compiles both course templates without local setup.

Evaluation

Assignments & grading

Grade Components

36%
Graded on completion, not correctness: a genuine attempt earns full credit. Written feedback on all work
30%
Teams of 4–5. Three check-ins: proposal, mid-point check-in, final presentation
18%
Two per student, 9% each. A standalone textbook chapter on one lecture, in the course LaTeX template
11%
Every class from Lecture 2, on paper, at a random time; handed to a TF. No make-ups
5%
Scored every class: 5 / 2.5 / 0. Average with the two lowest dropped

Grading Scale (YSPH)

≥ 90%Honors (H)
80–89%High Pass (HP)
65–79%Pass (P)
< 65%Fail (F)

Late Policy

Late submissions are not accepted. Contact the instructor before the deadline for documented emergencies.

Grading Turnaround

Homework & Scribe Notes1 week
ParticipationPosted at the end of the course

Participation is scored every class by the TFs and recorded by name; the term score appears on Canvas at the end of the course.

Key Dates

Per-lecture scribe deadlines are in the lecture calendar. Week 1 is Aug 31–Sep 4. Where this page and Canvas disagree, Canvas is right.

Course Project

Teams of 4 to 5, worth 30%; teams were set on Fri Sep 18. The proposal is due Tue Sep 29, 11:59pm. In one to two pages, describe your question and why it matters, the dataset you plan to use and how you will access it, a tentative analysis plan with a way to evaluate the results, and a timeline that shows each member’s responsibilities. No results are expected yet; if data access or the method is uncertain, say so and describe your next step. Use the proposal template (.tex · .pdf); the details are in the Canvas announcement. On Thu Oct 1, one speaker from each team presents the proposal in class for five minutes.

A list of potential project ideas (PDF), with collaborators and data links, is on this page; your own idea is welcome. Projects are judged on curiosity and scientific thinking, not novelty. Later check-ins: a mid-point check-in (week 8), the final presentation at a poster session in class (week 12), and a written report in finals week. See Key Dates.

Homework

Four assignments, 9% each. Points are for a genuine attempt with the reasoning shown, not for a correct answer; written feedback is given on all work. Each submission is three files on Canvas: hw#-lastname.pdf, .tex, and .py, written in the solution template (.tex · .pdf). No handwritten work, figures included.

Homework 2 is due Sat Oct 17 on Canvas. Homework 1 was due Sat Sep 26. Python and LaTeX setup instructions are in the Homework 1 handout; see the software stack.

Scribe Notes

Each student scribes two lectures, 9% each; assignments are on Canvas and change only for reasons such as illness. A scribe note is not lecture notes but a superset: a self-contained chapter on the lecture topic, with references, that goes beyond what was covered. Use the template (.tex · .pdf · .bib); read the examples for Lecture 1, Lecture 2 and Lecture 3. Notes become a shared resource for the class: graded student notes are posted next to their lecture in the calendar.

Submit both the .tex and the compiled .pdf. Deadlines by lecture are in the calendar.

In-Class Quizzes

Every class from Lecture 2 has a short paper quiz at a random time. Hand it to a TF before you leave. No AI, no devices, no make-ups. 11% total. Scores are posted to Canvas.

Participation

Scored by the TFs in every class. 5: substantive (guiding discussion, volunteering, reasoning through an answer, an objection, a problem at the board). 2.5: minimal (a clarifying question, a short answer, speaking only when called on). 0: none, or absent. Term score is the average with the two lowest dropped.

Bring your tent name card (template in Canvas Files) to every class; scores are recorded by name. The scribe for a lecture is recorded as fully participating. If speaking in class is a barrier, email the instructor in the first two weeks; written contributions can be credited instead. See also the device policy.

About This CourseCourse overview, prerequisites, and teaching team

Description

A comprehensive introduction to data science with a focus on public health and healthcare applications. Covers data structures and algorithms, machine learning, and causal inference — equipping students to decompose complex problems, select appropriate methods, and communicate results to diverse audiences.

Who Should Enroll

MS / PhD in Biostatistics Required
MPH students (quantitative focus)
Advanced undergrads (with permission)
Stats & DS or CS grad students

Instructor

Assistant Professor, Biostatistics
harsh.parikh@yale.edu

Office Hours

Thu, 10–11am

Teaching Fellows

Christopher Alvarez · Mon 11am–1pm, Zoom (link on Canvas)
Swaha Bhattacharya · Fri 1–2pm, hybrid · Kline Tower, 11th floor, or Zoom (link on Canvas)
Yukta Nagaraj · Tue 2–3pm, Zoom (link on Canvas)

Prerequisites

Foundational knowledge in the following areas, or concurrent enrollment in courses covering them.

Linear Algebra
Vectors, matrices, basic operations
Set Theory
Notation, union, intersection, complement
Probability
Random variables, distributions, Bayes' rule
Programming
Any language; Python helpful but not required

Learning ObjectivesWhat you will learn

By the end of this course, students will be able to:

1
Formulate
precise mathematical or statistical questions from natural language problems, identifying meaningful estimands
2
Decompose
complex problems into tractable sub-problems and recognize appropriate tools for each component
3
Implement
data science methods using Python (NumPy, Pandas, Scikit-learn) to analyze real-world datasets
4
Interpret
results within the appropriate problem context, understanding capabilities and limitations
5
Communicate
findings effectively to academic and non-technical audiences through writing and visualization
6
Evaluate
the appropriateness of analytical approaches, making informed trade-offs between methods

Course PoliciesAI use, collaboration, attendance, and submissions

AI Policy

AI assistants (ChatGPT, Claude, Copilot) are permitted as learning aids for homework and scribe notes:

  • You must understand everything you submit
  • Declare all AI use at the top: the tool, what it did, what you verified, what you did not (both templates have a declaration block)
  • Undeclared use, or an inaccuracy traced to AI → zero for the submission
  • You own all accuracy; auditing the output is your job
  • Quizzes: no AI allowed

Collaboration

Discussion with classmates is encouraged, but:

  • All written solutions must be your own
  • All code must be written individually
  • Acknowledge all collaborators and resources
  • Do not share answers, code, or LaTeX files
  • The course project is team work by design

Electronic Devices

Mobile phones, iPads, and laptops are not permitted during lectures without prior permission; take notes on paper. Unauthorized use can zero your participation grade for the whole semester. Permission is automatic with a Student Accessibility Services accommodation: email the instructor, no explanation to anyone else needed. No devices during quizzes under any circumstances.

Attendance

Attendance is mandatory. Every class has a quiz and a participation score, and neither can be made up. Bring your name card; participation is recorded by name. Class dates and recesses are in the calendar.

Submissions

All homework and scribe notes are submitted on Canvas, typeset in LaTeX from the course templates. No handwritten work. No late submissions. Check this site and Canvas announcements for updates.

Academic Integrity

All students must adhere to the YSPH Code of Academic and Professional Integrity (CAPI), in the YSPH student portal. Violations — plagiarism, unauthorized collaboration, undisclosed AI use, fabrication — will be referred to the CAPI Committee. Penalties can include expulsion.

ResourcesAccessibility, wellbeing, and university support

Accessibility

Yale Student Accessibility Services (SAS). Email sas@yale.edu or call 203-432-2324.

Mental Health

YSPH Wellness Counselor: Diane Frankel-Gramelis. 988 Lifeline: dial 988 (24/7).

Writing Support

Graduate Writing Lab — free consultations. Book at yale.mywconline.net.

Inclusivity

Office of Community & Practice. Contact Mayur Desai or Randi McCray.

Title IX

Deputy coordinator: Kelly Shay. Office of Institutional Equity & Accessibility.

Classroom Safety

emergency.yale.edu · Classroom preparedness · Academic calendar