project-2-final-jlyc-fp created by GitHub Classroom

Team members: Xiao Ji ([email protected], xj2247), Xinyi Liu ([email protected], xl2904), Jiaying Chen ([email protected], jc5299), Duanyue Yun ([email protected], dy2400)

Please find the Final Report.ipynb for our final report.

Repository contents:

• Bias baseline & user, item segmentation.ipynb: bias baseline model, user and item segmentation, build user attributes
• CMF.ipynb: code for collective matrix factorization
• Business.ipynb: exploratory analysis on business.json
• MF baseline - ALS.ipynb: Matrix Factorization (using Spark ALS) baseline model
• lightFM - Feature selection.ipynb: Factorization machine model feature selection
• lightFM - cross validation.ipynb: Factorization machine model hyper-parameter cross validation
• lightFM full dataset - overall results.ipynb: Factorization machine model full dataset precision and AUC
• lightFM full dataset - precision by segment.ipynb: Factorization machine model full dataset precision by active/moderate/non-active users and popular/moderate/unpopular items
• lightFM full dataset - auc by segment.ipynb: Factorization machine model full dataset AUC by active/moderate/non-active users and popular/moderate/unpopular items