Time Series Models: Approaches to Forecasting : A Tutorial
Published on: Jan, 25, 2011
Time Series Models
- Quantitative forecasting models that use chronologically arranged data to develop forecasts.
- Assume that what happened in the past is a good starting point for predicting what will happen in the future.
- These models can be designed to account for:
- Seasonality effects
- Can quickly be applied to a large number of products
- Forecast accuracy measures can be used to identify forecasts that need adjustment (management by exception
Randomness & trend
Randomness, trend & seasonality
h2. Basic Idea Behind Time Series Models
…Distinguish between random fluctuations & true changes in underlying demand patterns.
Simplicity is a virtue – Choose the simplest model that does the jobh2. Moving Average Models
- Based on last x periods
- Smoothes out random fluctuations
- Different weights can be applied to past observations, if desired
h2. Table of Forecasts & Demand Values…
h2. ... and Resulting Graph
Note how the forecasts smooth out variations.
Read the Supply Chain Management Professional Newsletter
Read the latest supply chain research, articles, and news as soon as we post them.