SCM 303 Lecture Notes - Lecture 8: Delphi Method, Time Series
Document Summary
What is forecasting: forecasting predicts the future, usually demand volumes (downstream, also used for volume and supply prices (upstream) Critical information for planning: production capacity, supplier contracts, distribution capacity, new products, revenues. Laws of forecasting: forecasts are always wrong, forecasts tend to be more accurate for: Near term vs. long term time periods. Types of forecasts: qualitative (subjective opinion, quantitative (math based) Qualitative methods: market survey, panel consensus. Delphi method: poll a panel of experts individually on what they think the forecast will be: life cycle analogy, build-up. Time series analysis types: average, moving average, weighted moving average, exponential smoothing, double exponential smoothing. Always plot data points to make the best decision on what forecasting model to choose. Casual models: time series uses pas demand data to predict future. How good is my forecast: after demand is realized, compare actual demand to the forecasted demand.
