Regression Analysis-Based Predictive Model for E-Commerce Application

Online sales have been growing every year. It is necessary for the seller to learn about the market choice to make a profit. The black Friday Sales dataset from Kaggle is used to train the prediction model. We get the purchasing power of the customer from the dataset. The prediction model is done using regression algorithms like Linear regression, Ridge regression, and Lasso regression and also using regression trees like Random forest, and Decision tree regressor. The performance evaluation is done using performance metrics like Mean Square Error(MSE), Mean Absolute Error(MAE), and Root Mean Square Error(RMSE). The model which uses the Random Forest Regressor outclasses the other training models with the least MAE, and MSE scores of 47.19 and 3062.72. Also the Decision Tree Regressor performed well with the MAE, and MSE scores of 48.73 and 3363.87, with better timing of 16s, whereas Random Forest Regressor took 6 mins. The Random Forest Regressor has the better Cross-Validation score of 3053.05 followed by the Decision tree with a score of 3340.25.