Theoretical Machine Learning Courses and Certifications

Information Theory of Deep Learning - Naftali Tishby

Explains deep neural network learning through the Information Bottleneck, linking input compression and mutual information to generalization and accuracy.

• YouTube
• 1 hour 25 minutes
• Self-Paced
• Free Video

Understanding Artificial Intelligence through Algorithmic Information Theory

6 ratings at edX

Can we characterize intelligent behavior? Are there theoretical foundations on which Artificial Intelligence can be grounded? This course on Algorithmic Information will offer you such a theoretical f

• edX
• 5 weeks, 4-8 hours a week
• Self-Paced
• Free Online Course (Audit)

Foundations of Machine Learning

Explore foundational machine learning through formal theory, SGD optimization, overparameterized regression, and energy-based models connecting classical ideas to modern systems.

• Swayam
• Available until Oct 9, 2026
• Free Online Course

IIT Bombay;  NPTEL

Bandit Algorithm (Online Machine Learning)

Study bandit algorithms for balancing exploration and exploitation in stochastic and adversarial online learning settings.

• Swayam
• 12 weeks
• Available until Oct 9, 2026
• Free Online Course

Google TechTalks

Representational Strengths and Limitations of Transformers

Explore the mathematical analysis of attention layers in transformers, comparing their representation power to other architectures and examining complexity parameters in various tasks.

• YouTube
• 58 minutes
• Self-Paced
• Free Video

Serrano.Academy

The Kolmogorov-Arnold Theorem - Understanding Continuous Functions and Neural Networks

Discover how continuous functions can be expressed through simple formulas using the Kolmogorov-Arnold Theorem and its practical applications in neural networks.

• YouTube
• 23 minutes
• Self-Paced
• Free Video

Metrized Deep Learning - Lecture 23

Explore Jeremy Bernstein's research on metrized deep learning, covering modular neural network design, optimization theory, scaling principles, and duality concepts.

• YouTube
• 1 hour 8 minutes
• Self-Paced
• Free Video

Generalization Theory - Overparameterization, Double Descent, and Inductive Biases - Lecture 6

Explore overparameterization, double descent, and VC dimension limitations while understanding how inductive biases shape deep learning generalization.

• YouTube
• 1 hour 21 minutes
• Self-Paced
• Free Video

Google TechTalks

Tree Learning: Optimal Algorithms and Sample Complexity - Hierarchical Clustering and PAC Learning

Explore optimal algorithms and sample complexity for learning hierarchical tree representations from labeled data, covering PAC learning, online learning, and efficient tree classifier construction.

• YouTube
• 35 minutes
• Self-Paced
• Free Video

Google TechTalks

Streaming Attention Approximation via Discrepancy Theory

Discover BalanceKV, an algorithm using geometric properties to compress key-value cache in LLMs, reducing memory requirements while maintaining attention quality through discrepancy theory.

• YouTube
• 33 minutes
• Self-Paced
• Free Video

UofU Data Science

Online Learning and the Mistake Bound Model - Lecture 7

Explore the principles of online learning algorithms and performance quantification through the mistake bound model, focusing on theoretical foundations and practical applications.

• YouTube
• 1 hour 19 minutes
• Self-Paced
• Free Video