1. Classification with Hyperparameter Search : The idea here is to train and evaluate 8 classification methods across 10 classification datasets. 2. Regression with Hyperparameters Search: The idea here is to train and evaluate 7 regression methods across 10 regression datasets. 3. Classifier interpretability : load and train models on standard computer vision dataset called CIFAR-10 and train a convolutional neural network using PyTorch to classify images in the dataset; train a decision tree to classify images in the dataset; and try to interpret the CNN using the 'activation maximization' technique. 4. Novelty component : Try to introduce a novel aspect to your analysis of classifiers and regressors or to your investigation of interpretability.
