Overview
Dive into a comprehensive course on building neural networks from scratch without relying on libraries. Learn to manually adjust network parameters to teach a car how to navigate a custom-designed playground. Gain insights from Dr. Radu, a university computer science professor with over a decade of experience in machine learning and AI. Explore fundamental concepts through hands-on lessons covering topics such as single neurons, genetic algorithms, hidden layers, and Dijkstra's algorithm. Engage with homework assignments, participate in data collection through racing simulations, and compete for prizes. Access additional resources including code repositories, related tutorials, and community discussions to enhance your learning experience.
Syllabus
Use "11. MiniMap" from Phase 2 at
⌨️ Introduction
⌨️ The Playground
⌨️ One Neuron
⌨️ Clarrifications
⌨️ Lesson 2
⌨️ Genetic Algorithm
⌨️ 2 Inputs
⌨️ Hidden Layers
⌨️ Misconceptions
⌨️ Lesson 3 More Outputs
⌨️ Lesson 4 Traffic Rules
⌨️ Lesson 5 Compass Sensor
⌨️ The need for Shortest Path
⌨️ Updating the Self-driving Car codebase
⌨️ Lesson 6 Dijkstra's Algorithm
⌨️ Lesson 7 Dijkstra with AI Agents
⌨️ Final Challenge
Taught by
freeCodeCamp.org