Learn Neural Networks With Go - Not Math!
summary
Claims that neural nets are learnable with code not math and then implements neurons, bias, mean‑square error, gradients and backprop in real Go: Go only, no TensorFlow or PyTorch. Demo builds a single neuron that improves, a layer with sigmoid activation, and runs roughly 250 training cycles showing learning; uses Garden to rebuild and redeploy during edits. Code repository exists but the full multi‑layer gradient chaining is not finished.
AI-written summary · about
chapters (13)
- 0:00Intro
- 0:42What are Neural Networks
- 2:57What is a Neuron
- 4:06Bias
- 5:16Loss
- 5:43Mean Square
- 9:13Gradients
- 12:20Gradient Descent
- 17:15Mean Square Error
- 18:14Layers
- 19:27Layer Demo
- 21:55Sigmoid Function
- 27:36Prediction
related talks
29mLearn Neural Networks With Go, Not Math
06mLearn Neural Networks With Go—Not Math! (Interview)
50mDeep Learning with Go
51mMachine Learning Exposed Deep and Reinforcement Learning
43m(Deep) Learn You a Neural Net for Great Good!
44mMachine Learning Exposed
43mA Gentle Introduction to Neural Networks
37mCode First, Math Later: Learning Neural Nets Through Implementation and Examples