Our demand forecasting model's error dropped from 18% to 9% MAPE. The biggest improvement wasn't a fancier…
Priya Banerjee
Data Scientist | NLP, forecasting, Python | Ex-researcher | Women in Tech mentor
Our demand forecasting model's error dropped from 18% to 9% MAPE. The biggest improvement wasn't a fancier model. It was adding Indian festival calendars and local holidays as features. Diwali week looks nothing like a normal week 😄
21 reactions2 comments