Implementation of various deep learning models from scratch

Models are implemented from scratch in PyTorch using only tensor operations.

Models Implemented :

1. Logistic Regression (Not technically a deep learning model, but gives a base for Neural Networks)
2. Deep Neural Network (Fully-connected layers only)

Note : Only binary classification done till now. Will add multi-class classification also.

Optimization Algorithms Implemented :

1. Batch Gradient Descent
2. Mini-batch Gradient Descent
3. Stochastic Gradient Descent (put batch_size = 1 in mini-batch to get this)
4. Batch Gradient Descent with Momentum
5. Batch RMSProp
6. Batch Adam

Training and Validation results are added in the Results folder.