Example: Update weight B4 (x1→h1): =B4 - $Z$1 * W14 (assuming W14 holds gradient for that weight)
In the age of Python, TensorFlow, and PyTorch, it is easy to forget that the core of a neural network is just matrix multiplication, activation functions, and gradient descent. Surprisingly, you can build a fully functional, trainable neural network using nothing but native Excel formulas.
He wrote a tiny macro, a script to automate the learning loop:
), and the forward propagation formulas will instantly calculate a prediction based on your "learned" weights. Excel formula template