Skip to content

Bujiazi/DeepMovie

Repository files navigation

DeepMovie

Brief Introduction

Official implementation of the AI3602 course project DeepMovie developed by Group 6 Gold Miners. It is a plug-and-play deep nueral network training pipeline turning most deep learning base modules into a movie recommender system.

          

The brief process of DeepMovie can be summarized as the following graph:

          

Prepare Datasets

MovieLens-1M and 10M datasets can be downloaded from MovieLens. Please put your dataset under the data folder along with ml_plot.dat.

Data Preprocessing

python data_preprocess.py 

DNN Training and Evaluating

python train_evaluate.py 

Quantitative Results (in RMSE)

model movielens-1m movielens-10m
PMF(baseline) 0.8971 0.8311
CNN 0.8733 0.7970
CNN_KAN 0.8725 0.7941
LSTM 0.8675 0.7959
ResNet 0.8658 0.7931
Transformer 0.8601 0.7883

Contact Us

Jiazi Bu: [email protected]
Zhiyuan Zhang: [email protected]
Xizhuo Zhang: [email protected]

About

SJTU AI3602 Spring course project

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages