Machine Learning Classification using Adam Optimizer.
The code is written using numpy library to achieve the following:
- Randomized initial weights (lines 18-24).
- Created a sigmoid layer for forward pass (lines 34-42).
- Trained NN with Adam's optimizer.
- Sample input (lines 56-73). *Data could be obtained from http://yann.lecun.com/exdb/mnist/
- Stopping criteria (line 76 and 77).
- Customisable hidden layers numbers and nodes (line 78).