HarvardX: Data Science: Building Machine Learning Models course Syllabus

A rigorous and concept-driven course that builds a strong foundation in machine learning for data science. This course spans approximately 9–12 weeks with a weekly commitment of 6–8 hours, combining theory, hands-on practice, and real-world applications. Learners will progress through core machine learning concepts, from supervised and unsupervised learning to model evaluation and practical implementation, culminating in a final project that integrates all learned skills.

Module 1: Introduction to Machine Learning

Estimated time: 10 hours

• Understand what machine learning is and how it fits into data science
• Distinguish between prediction and inference
• Explore real-world applications of machine learning
• Identify types of machine learning problems

Module 2: Supervised Learning Methods

Estimated time: 16 hours

• Learn linear regression and logistic regression fundamentals
• Understand classification basics
• Work with training data and labels
• Evaluate prediction accuracy

Module 3: Unsupervised Learning and Clustering

Estimated time: 16 hours

• Apply k-means clustering techniques
• Discover patterns in unlabeled data
• Understand dimensionality reduction concepts

Module 4: Model Evaluation and Validation

Estimated time: 16 hours

• Implement cross-validation and resampling techniques
• Evaluate models using appropriate performance metrics
• Understand overfitting, underfitting, and the bias–variance trade-off
• Select models that generalize well to new data

Module 5: Practical Machine Learning Applications

Estimated time: 16 hours

• Apply machine learning workflows to real-world datasets
• Interpret model outputs and limitations
• Understand ethical considerations and responsible use of ML models

Module 6: Final Project

Estimated time: 20 hours

• Build and train a machine learning model using real data
• Evaluate model performance with appropriate metrics
• Submit a report interpreting results and ethical implications

Prerequisites

• Basic understanding of statistics and probability
• Familiarity with Python programming
• Introductory knowledge of data analysis concepts

What You'll Be Able to Do After

• Understand core concepts behind modern machine learning in data science
• Apply classification, regression, and clustering techniques to real-world datasets
• Build and evaluate supervised and unsupervised learning models
• Choose appropriate machine learning approaches for different problems
• Interpret model performance and make data-driven decisions