Regression using Optimization techniques

I have calculated Mean Square Error using Sigmoid neuron and Feed Forward Network for my input data which has 3 input columns and 1 output.
I have also been able to implement backward propagation on my data and create the heat map for w and b.

In this series, I want to implement various optimization techniques as taught in 0404 and 0407 dated files GDAlgorithms.
I tried but getting stuck as to what values to take for initiation w_int and b_int.

Please give more insights regarding this.

image

Hi @nishantra4,
Maybe you can try out Xavier or He initializations,