AI Foundations: From Math to Neural Networks
Deconstruct how neural networks compute using fundamental math. Focuses strictly on conceptual mechanics rather than training pipelines or engineering.
Module 1The Forward Pass
Understand how data flows forward through a neural network—from feature vectors and dot products to artificial neurons and activation functions.
The Forward Pass
Understand how data flows forward through a neural network—from feature vectors and dot products to artificial neurons and activation functions.
01 • Linear Algebra Part 1 (Vectors & Dot Products)
The Vector (State & Identity)
Learn how real-world data is structured into ordered numeric lists and vector coordinates to create mathematical identity profiles in AI systems.
Combining Vectors (Addition & Scaling)
Master vector addition and scalar multiplication to transform feature states, scale attributes, and combine multidimensional data profiles.
The Dot Product: The Computational Engine
Learn how the dot product multiplies input features by weights to accumulate influence, calculate raw decision scores, and power neural network layers.
The Big Picture: From Math to Neural Networks
Connect feature vectors and dot products directly to neuron weights, biases, and activation thresholds in modern deep learning architectures.
Do It By Hand
Solidify your understanding with hands-on practice problems, calculating vector arithmetic, projections, and dot products manually.
The Backward Pass & Optimization
Master loss functions, partial derivatives, the multivariate chain rule, and backpropagation mechanics to train and optimize neural network parameters.
Synthesis & Modern AI Foundations
Synthesize the forward and backward pass to understand multi-layer networks, deep learning architectures, and modern transformer foundations.