Statistical models have been applied in many fields. In most cases, the models are built based on observational data, which are usually not as good as experimental data, which are obtained through statistical experimental design (DoE), such as full factorial design, response surface design, etc.
This project will show how to obtain reliable model performance results through good experimental design, and how to transform a complex linear prediction model into a practical web app. This case can also be applied to other machine parameter learning research.
