Domain Driven Forecasting: Applying Theory Into Practice

摘要

In this paper, we present our experiences in applying state-of-the-art forecasting solutions to meet the forecasting needs of various business domains. We present four real-world case-studies varying in the business objectives, forecasting needs, and domain properties. Each case-study presented unique challenges in translating theory into practice and translating forecasting observations into domain-specific recommendations. We summarize the lessons learnt while deploying across various case-studies, and demonstrate how the state-of-the-art solutions coupled with industry best practices can deliver powerful solutions to meet forecasting needs of any business domain.