PyTorch-Classification-and-Regression

This repository includes a classification training on MNIST dataset and regression on California Housing dataset using neural networks with PyTorch.

The aim of this project is to try and find the best combination of optimizer, activation and loss function, network model, and different training modes (batch, mini-batch, stochastic).

This notebook includes 2 projects.

1. A neural network training model for classification on MNIST dataset.
2. A neural network training model for regression on California Housing dataset.

The outputs for this project are as follows:

1. Model: MNISTModel2, Optimizer: SGD, Loss Function: MultiMarginLoss, Batch Size: 64, with Validation Accuracy: 0.9343
2. Model: RegressionModel1, Optimizer: Adam, Loss function: SmoothL1Loss, Batch size: 1, Test loss: 0.7584