Task $0$: Preparation

perform. Please note that you should use each dataset's original train$/$test splits to train and report the test scores.

Task $1$: Time series classification using deep learning $1$

following rules:

When initially loading the dataset, use sktime's load$\_$UCR$\_$UEA$\_$dataset function. This is for our grading purpose.

You should use at least two Tensorflow callbacks when you fit your model. These can be built$-$in ones or your personalized callback.

You need to clearly report the test accuracy of the four models. Training and validation accuracy scores are not enough.