1.implement the neural network model on a shale dataset 2.find which parameters are the key perfomance indicators 3.generate type curves for production 4.do prediction for production based on simulative data 5.finish the sensitivity analysis for all parameters