English | MP4 | AVC 1280×720 | AAC 48KHz 2ch | 1h 05m | 173 MB
Exponential smoothing is a term for a set of straightforward forecasting procedures that apply self-correction. Each forecast comprises two components. It’s a weighted average of the prior forecast, plus an adjustment that would have made the prior forecast more accurate. Smoothing—like most credible approaches to forecasting—requires a baseline of observations, in sequence, to work properly. Weekly revenues and daily hospital admissions are typical examples. Several versions of exponential smoothing exist, each corresponding to a type of baseline. In this course, Conrad Carlberg provides an introduction to simple exponential smoothing, diving into the basic idea behind it, and explaining how to assemble the forecast equation and optimize forecasts.
Topics include:
- Using correlograms to identify the nature of a baseline
- Assembling the forecast equation
- Methods of identifying the first forecast
- Getting a measure of overall forecast accuracy
- Optimizing a smoothing constant by minimizing RMSE
Table of Contents
Introduction
Welcome
The Idea Behind Exponential Smoothing
Exponentially declining influence of observations
From error correction to smoothing
Identify the appropriate baseline
Self-correcting forecasts
The Forecasting Equation
Dissect the error correction form
Dissect the smoothing form
Exponential smoothing tool
Initialize the forecasts
Measuring Forecast Accuracy
The absolute deviation approach
The least squares approach
Optimizing Forecasts
Set up the smoothing formula for Solver
Solution in R
Solver
Conclusion
Next steps
Resolve the captcha to access the links!