Xavier Initialization
A mathematical formula for setting the initial random weights of a neural network to ensure signal variance remains perfectly constant across all layers.
Think of It Like This
Like tuning every instrument in an orchestra to the exact same base volume before the concert starts so no section drowns out the others.
Also known as Glorot initialization, it draws initial weights from a distribution scaled precisely by the number of input and output connections (fan-in and fan-out) of a neuron. It prevents activations from exponentially exploding or vanishing to zero during the critical first few epochs of training, stabilizing deep learning.