STEM Program

Fundamental AI and Machine Learning

An AI Engineer Training Program with Live Python Tutorials

with Harvard Professor Dr. Xing · 96 Instruction Hours

Program Overview

From zero experience to a working ML engineer's toolkit.

This program is an intensive, hands-on introduction to applied machine learning and AI engineering, designed for beginners. It emphasizes practical skills, intuitive understanding, and real-world problem solving.

By the end, students are capable of performing the core tasks expected of a junior AI or machine learning engineer, including data analysis, model development, evaluation, and basic optimization.

01
Foundations
Learn programming and core AI concepts
02
Modeling
Construct and understand machine learning models
03
Evaluation
Assess performance and test outcomes
04
Application
Solve real-world data challenges in healthcare, finance, and more
Faculty Advisor

Dr. Fangxu Xing

Professor, Harvard University
15+ years of experience in AI, machine learning, and computer vision
11 years at Harvard; trained 800+ students
Authored hundreds of peer-reviewed publications

Dr. Fangxu Xing is a Harvard professor with over 15 years of experience in artificial intelligence, machine learning, and computer vision. He earned his degree from Johns Hopkins University, trained as a postdoctoral researcher and AI engineer at Massachusetts General Hospital, and has since authored hundreds of peer-reviewed publications on AI applications in healthcare and beyond.

Over 11 years at Harvard, he has hands-on trained more than 800 students and learners, many of whom have gone on to major positions in industry, including FAANG+ companies.

Passionate about AI education, Dr. Xing makes machine learning approachable, practical, and engaging, helping students connect core theory to real-world impact.

Selected Publications
  • Free-mask: A novel paradigm of integration between the segmentation diffusion model and image editing
  • Label Space-Induced Pseudo Label Refinement for Multi-Source Black-Box Domain Adaptation
  • Deep unsupervised domain adaptation: A review of recent advances and perspectives
  • Adversarial unsupervised domain adaptation with conditional and label shift: Infer, align and iterate
  • Subtype-aware unsupervised domain adaptation for medical diagnosis
Program Detail
Format
Online, live
Duration
12 weeks
Time commitment
6–8 hrs / week
Cohort size
20–30 students
36hours with the Professor
12Professor office hours
48Education Team office hours
96 instruction hours in total, including writing support during Education Team office hours.
Program Highlights
In depth
Significant contact time, 48+ hours, and support from the Faculty Advisor.
Beginner-friendly
Concepts introduced intuitively before the math behind them.
Hands-on first
Every topic reinforced with coding practice.
Case-driven
Real-world datasets and industry-inspired problems.
Engineering mindset
Focus on building usable, testable AI systems.
Progressive difficulty
From simple models to full ML pipelines.
What To Expect Weekly

Every session runs on the same loop.

Roughly three hours, split into two to three algorithm modules. Each module moves through the same four stages before a short break and the next module begins.

LEARN THE THEORY Slides + live handwritten derivations SEE IT IN ACTION Live Python demo on Colab CHECK UNDERSTANDING Quick quiz, then open Q&A RECHARGE 5–10 min break, then repeat ~3 hrs per weekly session
Program Structure

Twelve weeks, five stages.

Combining lectures, live coding, case studies, and hands-on projects, building from Python fundamentals to a full capstone project.

Weeks 0–2
Foundations
  • Introduction to AI and machine learning
  • Python programming for AI
  • Data structures, scientific libraries (NumPy, Pandas, Matplotlib)
  • First classification and regression models (KNN, Naive Bayes, Linear & Logistic Regression)
Weeks 3–4
Core ML Models
  • Decision Trees and Support Vector Machines
  • Bias–variance tradeoff
  • Ensemble methods: bagging, boosting, stacking
  • Introductory neural networks
Weeks 5–6
Math & Optimization
  • Intuitive calculus for machine learning
  • Gradient descent and optimization algorithms
  • Statistics for ML: correlation, Gaussian models, Central Limit Theorem
Weeks 7–9
Data Skills & Unsupervised Learning
  • Exploratory data analysis and data cleaning
  • Feature engineering techniques
  • Dimensionality reduction (PCA, t-SNE)
  • Clustering methods (K-Means, Hierarchical, DBSCAN)
Weeks 10–12
Evaluation & Final Project
  • Model evaluation metrics (accuracy, precision, recall, F1, ROC/AUC)
  • Regularization and hyperparameter tuning
  • Capstone project using a real-world dataset
What You'll Walk Away With
Program Outcomes
  • Weekly coding exercises
  • A final end-to-end AI project
  • A written technical report or paper
  • A formal program evaluation
  • A potential letter of recommendation from a Harvard professor
Skills You'll Build
  • Build and evaluate machine learning models independently
  • Understand how AI systems work in real-world applications
  • Read and modify existing ML codebases
  • Complete an end-to-end AI project using real industry datasets
Career & Academic Pathways
  • Build a strong foundation for advanced study in deep learning, computer vision, NLP, and AI research
  • Prepare for internships and entry-level roles: Junior AI Engineer, Junior ML Engineer, ML-focused Data Analyst, AI Research Assistant
Who Should Enroll

High schoolers and undergrads, starting from zero.

This program is open to high school and undergraduate students interested in AI engineering, data science, or applied AI careers, who have no prior experience.

Prerequisite: basic algebra only
How To Apply
1
Submit intake form & video
Fill out the intake form on our website. We also encourage a 1–3 minute video introducing yourself.
2
Online interview
Shortlisted applicants are invited to a 15-minute online interview with our admissions team.
3
Admission
Admitted students sign their program agreement and confirm enrollment by making payment. Then the journey begins.

Ready to start building?

Space is limited to 20–30 students per cohort. Apply now to secure your seat this fall.

Apply Now