Implementation of neural network (classification and regression)
- For both types sigmoid activation function was used
- For classification log loss was used with softmax layer for the last layer
- For regression quadratic loss was used
Implementation of neural network from scratch (classification and regression)
PythonMIT