We explore the effects of increasing the training dataset size, adjusting model complexity, introducing noise to the data, and employing weight decay regularization during training.

Classification using Softmax, Multilayer Perceptron (MLP) and Convolutional Neural Network (CNN) classifiers, comparing accuracy and loss performance in their training and testing phases.

Classification using Vanilla-RNN and LSTM.

Using Variational AutoEncoder (VAE) and Generative Adversarial Network (GAN).

Using Variational AutoEncoder (VAE) and Generative Adversarial Network (GAN).