The project aims to predict the first destination of a new customer to enable better targeting of offers and discounts. Thorough EDA was done to identify features that could probably impact the selection of destination. These features were then used in prediction model i.e. Random Forest and KNN to forecast the first destination of a new customer.

The project deals with forecasting sales at Store-Department-Week level. The technique used in this project includes Time Series ARIMA, Ridge Regression and Boosting.