BMO 2201 Lecture Notes - Lecture 10: Mean Absolute Percentage Error, Fourier Analysis, Process Design
Document Summary
Description of forecast management process components: forecast database must include timely historical and planning information. Must facilitate data manipulation, summarization, analysis and reporting. E. g. , open orders, demand history, marketing tactics, economy, competitor actions: technique is the computational method used to combine model components into a forecast quantity. E. g. , time-series or correlation modeling: support system must facilitate the maintenance, updating and manipulation of the database and the forecast, administration includes organizational, procedural, motivational, cross-functional and personnel aspects of forecasting. Meaningful forecast process requires integrated and consistent combination of components: faulty communications are costly for supply chains. Seek to reduce forecast inconsistency across multiple members of the supply chain: efforts to perfect a single component do not overcome need for other components, process design should consider strengths and weaknesses of each individual component. Design for optimal performance of integrated system. Bullwhip effect showing requirements error amplification between supply chain partners.