Question: A two-layer network with a single input, x, a sigmoidal hidden neuron, h, and a linear output neuron, o, has the following initial weights and

A two-layer network with a single input, x, a sigmoidal hidden neuron, h, and a linear output neuron, o, has the following initial weights and biases: Whx = 1, bn = 1, Woh = -2, b. = 1; An input/target pair is given to be (x = 1, t = 1). Find the squared error e? Find de? /awhx at the initial weights and biases P.S: bh and be represent the biases for the hidden neuron h and and the output neuron o. The error e is equal to (t-o) and dez/awhx de/awhx = 2e *
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
