This project is intended to solve modified MNIST dataset by using traditional and deep learning approches. Here traditional methods we use SVM and KNN, deep learning methods we use CNN and RNN(LSTM). Besides, we combined CNN and RNN, CNN and SVM ,which recieve better performance than single model. In addition, we tried Min-max Module network by MLP to solve this problem, which also achieve better accuracy than single MLP model.
mg2015started/ML_final_project
This project is intended to solve modified MNIST dataset by using traditional and deep learning approches
Python