GitXplorerGitXplorer
m

talk-multiseasonal-2020

public
0 stars
0 forks
0 issues

Commits

List of commits on branch main.
Unverified
7d4cb6a4e8d38422d82c7263100b433ef3b097cd

rm cache

mmitchelloharawild committed 3 years ago
Unverified
0b50178ee00c2ddf8c5d82c89f0ff009ba2de1f6

Final draft

mmitchelloharawild committed 4 years ago
Unverified
4d9548d44fe4c9b800bb78784d4d3369b5be3137

Add modelling resources

mmitchelloharawild committed 4 years ago
Unverified
7493dd1c53f69553e4fa39a26c681b1b49cd0015

Finished data preparation and visualisation

mmitchelloharawild committed 4 years ago
Unverified
7c27f06e26ef04fb4a794126e3169e33199a6ac1

Refined intro comparisons of models

mmitchelloharawild committed 4 years ago
Unverified
3510a5b026c23812bd42db346eec4c3a703a7687

Added intro to common seasonal terms

mmitchelloharawild committed 4 years ago

README

The README file for this repository.

Forecasting with multiple seasonality

NHS webinar Duration: 90 minutes. 60 minute talk with questions.

Slides and resources for multiple seasonal forecasting seminar given to NHS in 2020.

Abstract

Improved data collection procedures has provides more potential for time series analysis and modelling. Many seasonal time series models have been designed with monthly and quarterly data in mind, where only a single seasonal pattern is evident. For time series, collecting data at higher frequencies (or possibly even measuring event times) reveals multiple seasonal patterns that require more flexible models to handle. This webinar will discuss the challenges that multiple seasonality presents, and introduce some strategies used to model and forecast these types of time series.