Unsupervised Learning, Recommenders, Reinforcement Learning Course Syllabus

Overview: This course builds on foundational supervised learning knowledge, guiding learners through core unsupervised learning techniques and an introduction to reinforcement learning. With a focus on real-world applications, you’ll explore clustering, dimensionality reduction, recommender systems, and decision-making models. The course spans approximately 4 weeks, with 2–3 hours of study per module, combining theory and hands-on programming exercises. By the end, you’ll implement key algorithms in Python and complete a final project integrating multiple techniques. Lifetime access ensures flexible, self-paced learning.

Module 1: Clustering & k-means

Estimated time: 7 hours

• k-means clustering algorithm
• Elbow method for cluster selection
• Choosing the optimal number of clusters
• Hands-on: Clustering image data and customer segments

Module 2: PCA (Principal Component Analysis)

Estimated time: 7 hours

• Dimensionality reduction concepts
• Variance explained and principal components
• PCA implementation in practice
• Hands-on: Compressing and visualizing high-dimensional data

Module 3: Recommender Systems

Estimated time: 7 hours

• Content-based filtering
• Collaborative filtering methods
• Low-rank matrix factorization
• Hands-on: Building a movie recommender system with real data

Module 4: Reinforcement Learning

Estimated time: 7 hours

• Markov Decision Processes (MDPs)
• Bellman equations and value functions
• Q-learning algorithm
• Hands-on: Applying Q-learning in game-like environments

Module 5: Real-World Applications of Unsupervised Learning

Estimated time: 6 hours

• Unsupervised learning in search engines
• Video recommendation systems
• Pattern discovery in customer behavior
• Case studies from retail and online platforms

Module 6: Final Project

Estimated time: 8 hours

• Implement a clustering solution on real-world data
• Build a simple recommender system using matrix factorization
• Apply reinforcement learning to a simulated decision-making task

Prerequisites

• Basic understanding of supervised learning (regression, classification)
• Familiarity with Python programming and data libraries (NumPy, pandas)
• Fundamental knowledge of linear algebra and probability

What You'll Be Able to Do After

• Apply k-means clustering to segment customers and analyze images
• Use PCA for dimensionality reduction and data visualization
• Build and evaluate recommender systems using real datasets
• Implement Q-learning in reinforcement learning environments
• Understand how unsupervised and reinforcement learning drive AI in industry applications