- Goal: Predict the amount of purchase for consumers
- Data source: please refer to the dataset in the repository
- Model: Ensemble Model(Stacking)/ Deep Neural Network/ Support Vector Regressor/ ElasticNet/ Regression Tree/ KNN Regressor
In this project, I first created a function to compiled the normalization and gridsearch into pipeline. The return of the function including the best gridsearch score, best parameter of the model, and the MSE of test set of each model. The result shows that Deep Neural Network is the best model with the lowest MSE score.