Exponential smoothing is a time series forecasting method for univariate data that can be extended to support data with a systematic trend or seasonal component.<\/p>\n
It is common practice to use an optimization process to find the model hyperparameters that result in the exponential smoothing model with the best performance for a given time series dataset. This practice applies only to the coefficients used by the model to describe the exponential structure of the level, trend, and seasonality.<\/p>\n
It is also possible to automatically optimize other hyperparameters of an exponential smoothing model, such as whether or not to model the trend and seasonal component and if so, whether to model them using an additive or multiplicative method.<\/p>\n
In this tutorial, you will discover how to develop a framework for grid searching all of the exponential smoothing model hyperparameters for univariate time series forecasting.<\/p>\n
After completing this tutorial, you will know:<\/p>\n
- \n
- How to develop a framework for grid searching ETS models from scratch using walk-forward validation.<\/li>\n
- How to grid search ETS model hyperparameters for daily time series data for female births.<\/li>\n
- How to grid search ETS model hyperparameters for monthly time series data for shampoo sales, car sales, and temperature.<\/li>\n<\/ul>\n
Let\u2019s get started.<\/p>\n
![\"How](\"https:\/\/machinelearningmastery.com\/wp-content\/uploads\/2018\/10\/How-to-Grid-Search-Triple-Exponential-Smoothing-for-Time-Series-Forecasting-in-Python.jpg\")
<\/p>\nHow to Grid Search Triple Exponential Smoothing for Time Series Forecasting in Python
Photo by john mcsporran<\/a>, some rights reserved.<\/p>\n<\/div>\nTutorial Overview<\/h2>\n
This tutorial is divided into six parts; they are:<\/p>\n
- \n
- Exponential Smoothing for Time Series Forecasting<\/li>\n
- Develop a Grid Search Framework<\/li>\n
- Case Study 1: No Trend or Seasonality<\/li>\n
- Case Study 2: Trend<\/li>\n
- Case Study 3: Seasonality<\/li>\n
- Case Study 4: Trend and Seasonality<\/li>\n<\/ol>\n
Exponential Smoothing for Time Series Forecasting<\/h2>\n
Exponential smoothing is a time series forecasting method for univariate data.<\/p>\n
Time series methods like the Box-Jenkins ARIMA family of methods develop a model where the prediction is a weighted linear sum of recent past observations or lags.<\/p>\n
Exponential smoothing forecasting methods are similar in that a prediction is a weighted sum of past observations, but the model explicitly uses an exponentially decreasing weight for past observations.<\/p>\n
Specifically, past observations are weighted with a geometrically decreasing ratio.<\/p>\n
\n
Forecasts produced using exponential smoothing methods are weighted averages of past observations, with the weights decaying exponentially as the observations get older. In other words, the more recent the observation, the higher the associated weight.<\/p>\n<\/blockquote>\n