Sales Forecasting Model
Data Mining and Text Mining Challenge 2020
The aim of this project is to predict the sales from a company of the current week. Among data available, for each week is provided the sales of the previous week, product weight, brand, price, volume on promotion, the unique identifier for the products and other ones.
The procedure followed to get an accurate estimator includes data exploration to get used to the different features provided, evaluation of different models and, finally, built of the model regressor. The models evaluated include Linear Regression, k-Nearest Neighbors, Random Forest Regressor and ARMA models.