Suppose there exists a two layered network $($one input layer, one hidden layer and one output layer$)$ where in there are $150$ nodes in the input layer, $30$ nodes in the hidden layer and $1$ node in the output layer to solve a regression problem. a$.$ Without using backpropagation method, can you specify a methodology$/$procedure to find out the weights of the network given $$N$$ training examples? Note: You may make use of elementary calculus to solve this problem. b$.$ For the above mentioned network architecture and given the same training data set, suppose two students would like to implement the back propagation algorithm and subsequent steps to find out weights of the network. Would both of the implementations are guaranteed to result in the same weights of the network. c$.$ If so substantiate your answer by giving appropriate reasons. Otherwise mention at least couple of steps in two implementation that would be responsible to get different weights