A human-robot dance duet. Source: TED Talk ted.com

The research project analyzes different human (demand planners) and machine (machine learning forecasting algorithms) decision-making structures: full-human to AI delegation, hybrid AI-to-human or human-to-AI, and aggregated human-AI. The main goal is to understand the best combinations to improve the demand forecast accuracy. Based on demand planning processes between an FMCG company and JD.com, an experimental research methodology is applied to more than 1000 SKUs. The results show how better design and implement efficient machine learning forecasting algorithms.

Read more.