Rooftop joins Kaggle competition – to determine if using machine learning can improve demand forecasts

Today Rooftop Analytics joined the Kaggle competition M5 Forecasting - accuracy and competition M5 Forecasting - uncertainty

Thereby we dedicate our time to identify an algorithm – applying machine learning which predicts the sales units and the related uncertainty of Walmart’s sales forecast for selected sales units for a 28 days period. We join the competition with 4700 teams worldwide.


We do this because it can contribute to improve how retailers perform demand forecasts and help to minimize waste of food and other goods. Improvements in this area benefits not only retailers but also the globe.


It's not completely new to leverage artificial intelligence in demand forecast but to leverage our capabilities of time series analysis, causality analysis, and unique data science competences, we are confident that we can bring a better measure of certainty of retailer’s demand forecasts.


This is a great opportunity to learn and become wiser on the subject. A subject which has our attention already. 

Read our article Predictive Analytics - A Way to Prevent Food Waste.