Hybrid CNN-XGBoost Model for Video Popularity Prediction

CSE 4683 : Machine Learning and Soft Computing, Fall 2022, Emma Wade, Michelle Hardin, John Austin Reed

Data and Environment :

1. environment.yml : conda environment, to create environment based on yaml https://edcarp.github.io/introduction-to-conda-for-data-scientists/04-sharing-environments/index.html
2. Data available here: https://bitgrit.net/competition/11 and in project Canvas submission

Source Code :

1. file-prep.py : prepares training and testing files including one-hot encoding of categorical variables, cycling encoding of time variables, lasso regression of image features, and joining all variables. output needed to run cnn-xgboost.py and XGBOOST.ipynb
2. cnn-xgboost.py : hybrid model and CNN model
3. XGBOOST.ipynb : XGBoost model
4. ML_Plotting.ipynb : figure and comparisons scripts