+ 1

Who can help with backpropagation for a neural Network?

I want to implement a simple neural Network that can be Trained to learn a linear function. However i dont know how to calculate and Programm the backpropagation to adjust the weights. Can someone complete my code? https://code.sololearn.com/cVv975Cjod5b/?ref=app

6th May 2019, 5:39 PM
Daniel
2 Antworten
+ 1
First,you can use neural network in JS and python!
6th May 2019, 7:58 PM
CodeFu
CodeFu - avatar
+ 1
I found some issues with your code : 1) you should normalize data be for using them to get better results : You can see this where I used averages 2) you need just two weights since you have only two variables as input 3) to optimize weights , the following formula is used if you have only one output : Wi = Wi - (y1-y0)*xi*alpha You missed x with y in your code😅 https://code.sololearn.com/cAJC71HTgQ2S/?ref=app
6th May 2019, 8:10 PM
Cihad Jasem Alhaji
Cihad Jasem Alhaji - avatar