The Math Behind Machine Learning

Real ML, fully demystified.

Every concept in machine learning is a mathematical rule. We show you every rule, derive it, explain why it was chosen — and if the math is new to you, we teach you that too. Starting from high school algebra, ending with transformers.

Start the course

The Curriculum

Part 1

Framing the Problem

What does "learning" actually mean?

4 lessons

44 min

Part 2 Prereq: 0b-derivatives, 0c-vectors-matrices

Linear Regression

The simplest model

5 lessons

60 min

Part 3 Prereq: 0b-derivatives

Gradient Descent

How the machine learns

10 lessons

120 min

Part 4 Prereq: 0d-probability

Classification

Yes/no problems and decision boundaries

9 lessons

108 min

Part 5 Prereq: 0b-derivatives, 0c-vectors-matrices

Neural Networks

Stacking transformations

9 lessons

106 min

Part 6 Prereq: 0b-derivatives, 05-neural-networks

Backpropagation

The chain rule, applied

5 lessons

64 min

Part 7 Prereq: 01-framing, 03-gradient-descent, 06-backpropagation

Regularization

Preventing overfitting

8 lessons

90 min

Part 8 Prereq: 0c-vectors-matrices

Convolutional Networks

The math of spatial features

9 lessons

106 min

Part 9 Prereq: 0c-vectors-matrices, 04-classification

Attention & Transformers

The math behind modern AI

6 lessons

76 min

Part 10

Putting It Together

From math to real models

4 lessons

54 min

Part 11 Prereq: 03-gradient-descent

Advanced Optimization

Beyond vanilla gradient descent

9 lessons

108 min

Part 12 Prereq: 05-neural-networks, 06-backpropagation

Normalization & Initialization

Making deep networks trainable

7 lessons

82 min

Part 13 Prereq: 05-neural-networks, 06-backpropagation

Recurrent Networks

Learning from sequences

8 lessons

106 min

Part 14 Prereq: 0e-information-theory, 06-backpropagation, 04-classification

Generative Models

Learning to create, not just classify

10 lessons

140 min

Part 15 Prereq: 0d-probability, 0c-vectors-matrices

Unsupervised Learning

Finding structure without labels

10 lessons

126 min

Part 16 Prereq: 01-framing, 04-classification

Tree Methods & Ensembles

The algorithms that rule tabular data

7 lessons

90 min

Part 17 Prereq: 0c-vectors-matrices, 09-transformers

Embeddings & Representation Learning

Teaching networks what things mean

7 lessons

88 min

Part 18 Prereq: 05-neural-networks, 08-cnns, 09-transformers

Transfer Learning

Standing on the shoulders of pretrained giants

7 lessons

88 min

Part 19 Prereq: 04-classification, 07-regularization

Evaluation & Model Assessment

Measuring what actually matters

8 lessons

98 min

Part 20 Prereq: 09-transformers, 18-embeddings

Language Models

The math behind GPT and beyond

7 lessons

90 min

Part 21 Prereq: 08-cnns, 09-transformers

Advanced Architectures

ResNets, ViTs, and the design of scale

8 lessons

106 min

Part 22 Prereq: 0c-vectors-matrices, 04-classification

Kernel Methods & SVMs

Geometry of margins and feature spaces

6 lessons

78 min

Part 23 Prereq: 0d-probability, 05-neural-networks

Reinforcement Learning

Learning from rewards, not labels

7 lessons

96 min

Part 24 Prereq: 05-neural-networks, 16-tree-methods

Interpretability & Fairness

Understanding and auditing model decisions

7 lessons

84 min

Part 25 Prereq: 05-neural-networks, 0c-vectors-matrices

Graph Neural Networks

Learning on structured relational data

5 lessons

64 min

Part 0: Math Foundations

Never touched calculus? Start here. Comfortable with derivatives and matrices? Skip ahead.

Expand

Math Refresher

Functions & Notation

Reading math without fear

3 lessons

26 min

Math Refresher

Derivatives

Calculus from scratch

4 lessons

44 min

Math Refresher

Vectors & Matrices

The language of data

6 lessons

66 min

Math Refresher

Probability & Statistics

Reasoning under uncertainty

6 lessons

66 min

Math Refresher

Information Theory

The language of uncertainty and compression

4 lessons

48 min