-
Task 1: Implement the method from sklearn to implement classification. Sampled 100 instances out of each label (1,000 instances in total). Randomly divide the dataset with 80-20 ratio as training and testing set.
-
Task 2: Applied PyTorch package to implement the CNN algorithm, and trained the model using the training set generated in the last task and use your model to predict labels in the test set. The structure of CNN is defined as follows.
-Structure: Convolutional layer -> Max pooling layer -> Convolutional layer - > Max pooling layer -> Fully connected layer x2 -> Softmax layer
- Task 3: Performed Adaboost, Bagging and using two base classifiers respectively, including decision tree and logistic regression.
- PDF file contains the hand-written part of the question.