Neural Networks Demystified: Part 2
summary
Implementation of a small feedforward net in Python: 2 inputs, 3 hidden units, 1 output; weights initialized randomly and learned parameters separate from hyperparameters. Forward pass is done with matrix math (X, W1, Z2, A2, W2, Z3) and elementwise sigmoid, using numpy dot for speed and batch inputs. The code produces outputs but they are poor because weights are untrained; next step is gradient-based training to fit the toy sleep/study → test score mapping.
AI-written summary · about
related talks
01h 50mNeural Networks: Walkthrough
56mA Gentle Introduction to Neural Networks and making your own with Python
43m(Deep) Learn You a Neural Net for Great Good!
52mNeural Networks Walkthrough
01h 25mBuilding & Optimization of Neural Networks from Scratch with Numpy
58mNeural Networks: A Primer
40mNeural Networks: Where do I start?!?
36mNeural Networks