We have experiences in builting predictive models for applications in risk management, healthcare, marketing and retailing. Some of our prediction projects include
- Default risk prediction of P2P loans
- Sepsis prediction
- Freight shipping cost prediction
- Marketing campaign response rate prediction
- Churn rate prediction
- Revenue forecast
We primarily use R and AWS to develop predictive models. Our favoriate tools include caret, mlr, XGBoost, and MXNet.