This projects implements a DNN from zero.
This project is helpful for those who want to learn how to implement a Deep Neural Network (DNN) without using popular Deep/Machine Learning frameworks such as Tensorflow and PyTorch. All the stages: weight initialization, forward propagation, activation function derivatives, backpropagation, weight update, training, among others, are addressed and can be observed in the class DNN.
Please, change the DNN configuration using the nn_config variable (after main) and try to test the examples provided: regression example, binary classification, and multi-class classification. Comment/Uncomment the lines of code provided for each example according your needs.