Demand forecasting
Retail Sector
Customer
Leading DIY and construction products chain in Spain.
Description
Demand forecasting for more than 900 references, including products with very seasonal and sporadic sales.
Results
Predictive model of 2-month demand for more than 900 references, for 30 different shops. Good model performance measured in different ways (R2, RMSE, MAPE).
Technology
Data processing with Spark (aggregation of more than 1 billion records to calculate weekly sales per reference). Modelling carried out with scikit-learn and LightGBM.