The Enterprise Data Science and Analytics team at M&S designs models for diverse aspects of supply chain and the product lifecycle management. A significant hurdle we encounter is that the financial value of these models can only be measured through trials after substantial investments in data preparation and model training. In this session, we will divulge our strategies for enhancing a model’s probability of success in trials by examining the effectiveness of current decision-making processes, quantifying the impact of improved forecast accuracy and the optimisation of decision-making rules and parameters that utilise model forecasts. Furthermore, we will explore the selection of optimal metrics for profitable model training and the use of probabilistic forecasting to improve model profitability
Technical level: Technical practitioner
Session Length: 40 minutes
Supported by
