working_of_a_neuralnet.txt

# A simple 2 layer NN

Input -> FeedForward -> FeedForward -> Output _
                                               |
     BackProp   <-  BackPropogation <-  Loss <--








Loss can be computed using various loss functions :
The sum of squares error being one of the most basic ones

sumOfSs = sum(from i 0 to n)^2

Our goal in training is to find the best set of weights and biases that minimizes the loss function.

In order to adjust the weights and biases we need to know derivatives of the loss function wrt weights and biases