XGBoost - A Competitive Approach for Online Price Prediction

This study generates price prediction suggestions for a community-powered shopping application using product features, which is a recent topic of a Kaggle.com competition sponsored by Mercari, Inc. As Ebay acquired Canadian... [ view full abstract ]

This study generates price prediction suggestions for a community-powered shopping application using product features, which is a recent topic of a Kaggle.com competition sponsored by Mercari, Inc. As Ebay acquired Canadian data analysis firm Terapeak, the importance of using “big data” and machine learning to improve pricing decision-support in business has been rapidly increasing. It could be a great competitive advantage for companies or individual sellers having highly accurate pricing decision-support. The authors used R and Kernels to perform text analysis to generate features from unstructured product features, then used XGboost and Multiple Linear Regression to dynamically predict product price. After doing some exploratory data analysis, we created text features with above/below average prices for the most important features in the dataset. XGBoost was able to handle the over 2,000 brands data while Multiple Linear Regression was not. XGBoost achieved the best performance, with a 0.513 test set RMSLE.