data_mining_homework
MNIST dataset: hand-written digit classification by CNN and SVM
stock market prediction: using LSTM-model in Pytorch
CNN based hand written classification
train.py and test.ipynb is for CNN-based model.
details in relevant report report for CNN model.
SVM based hand written classification
SVM.ipynb and SVM.html is for SVM-based model.
details in relevant report report for SVM model.
LSTM based stock market prediction
benchmark.ipynb is the benchmark for stock market prediction and hyperparameters can be changed in class Args.
train.ipynb is the lstm model prediction with feature engineering and data preprocessing, whose hyperparameters can be changed in class Args.
details in relevant report report for stock market prediction.