Let W_ij represent the weight between node i at layer k and node j at layer (k − 1) of a given multilayer perceptron. The weight updation using the gradient descent method is given by

1. AW(t+1) = W(t) + α ∂E/∂W
2. BW(t+1) = W(t) − α ∂E/∂W
3. CW(t+1) = α ∂E/∂W
4. DW(t+1) = −α ∂E/∂W

Pick an option to check yourself, or open the answer below.

Show answer & explanation

Correct answer: B. W(t+1) = W(t) − α ∂E/∂W

Explanation

Gradient descent moves the weight against the error gradient: W(t + 1) = W(t) − α ∂E/∂W, with learning rate 0 ≤ α ≤ 1.