Instructor:

Cecil Bozarth , PhD
North Carolina State University
Author of "Introduction to Operations and Supply Chain Management," 2nd edition, Pearson, Prentice-Hall

 

SECTION Index

Introduction

Basic Rules of Forecasting

Time Series Models

   1. Exponential Smoothing

   2. Double Exponential Smoothing

Advanced Techniques

   1. Single Regression

   2. Multiple Regression

Qualitative Methods

Measuring Forecast Accuracy

Forecasting Strategies

   1. 3-Step Process

   2. Q2 Forecasting

Summary

References

 

 

Approaches to Forecasting : A Tutorial

Double Exponential Smoothing

 

What Is Double Exponential Smoothing?

Time Series with Trend: Double Exponential Smoothing

 

What Is Double Exponential Smoothing?

…like regular exponential smoothing, except includes a component to pick up trends.

top

 

Time Series with Trend: Double Exponential Smoothing

Formula 

• Ft            =          Unadjusted forecast (before trend)

• Tt            =          Estimated trend

• AFt          =          Trend-adjusted forecast                                                                                                                               

Ft            =          a* At-1           +  (1- a)  *   (Ft-1 + Tt-1)

Tt            =          b* (At-1-Ft-1)  +  (1- b)  *   Tt-1

AFt          =          Ft   +   Tt

 

To start, we assume no trend and set our “initial” forecast to Period 1 demand.

 

… We then calculate our forecast for Period 2.

 

… But Period 2 demand turns out to be 20.

What is the trend estimate for Period 3?

 

By Period 4, the model is starting to pick up on the trend …

 

…And after a few periods the model “locks on” to the correct trend value.

(Of course, this example is simplified to make the numbers clearer.)

 

top