Seasonal Persistence Forecasting With Python
Last Updated on August 28, 2019
It is common to use persistence or naive forecasts as a first-cut forecast on time series problems.
A better first-cut forecast on time series data with a seasonal component is to persist the observation for the same time in the previous season. This is called seasonal persistence.
In this tutorial, you will discover how to implement seasonal persistence for time series forecasting in Python.
After completing this tutorial, you will know:
- How to use point observations from prior seasons for a persistence forecast.
- How to use mean observations across a sliding window of prior seasons for a persistence forecast.
- How to apply and evaluate seasonal persistence on monthly and daily time series data.
Kick-start your project with my new book Time Series Forecasting With Python, including step-by-step tutorials and the Python source code files for all examples.
Let’s get started.
- Updated Apr/2019: Updated the links to datasets.
- Updated Aug/2019: Updated data loading to use new API.
Seasonal Persistence
It is critical to have a useful first-cut forecast on time series problems to provide a lower-bound on skill before moving on to more sophisticated methods.
This is to ensure we are
To finish reading, please visit source site