Axon 02: Machine Learning & Vision

Welcome to the Machine Learning & Vision Axon. This track develops deep learning from first principles—moving from multi-dimensional loss landscapes and gradient optimization to deep neural network layers, vector backpropagation, and real-time computer vision object detection.

Modules in this Axon

1. Loss Functions & Optimization

• Concept 01: Measuring Errors with Loss Functions (MSE & MAE)
• Concept 02: Cross-Entropy & Classification Loss
• Concept 03: Gradient Descent & Learning Rates

2. Neural Layers & Activation Functions

• Concept 01: Linear Layers (Weights, Biases & Dot Products)
• Concept 02: Non-Linear Activation Functions (ReLU, GELU, Sigmoid)

3. Vector Backpropagation Engine

• Concept 01: Computational Graphs & Vector Chain Rule
• Concept 02: Building an Autograd Engine in Pure Python

4. Computer Vision & Object Detection

• Concept 01: 2D Spatial Convolutions & Feature Maps
• Concept 02: Object Detection, Anchor Boxes & IoU