Forecasting with Multiple Seasonality

Nov 25, 2020 1:00 PM — 2:30 PM

With the advances in digital technologies, data is recorded more frequently in many sectors including healthcare. For instance, hospitals collect the arrival time for each admission that contains a detail level of temporal granularity. This often results in time series that exhibit multiple seasonality of different lengths.

Traditional approaches such as simple methods, exponential smoothing or ARIMA models are not appropriate for series with multiple seasonal cycles. Forecasting problems involving such series have been increasingly drawing the attention of researchers leading to the development of several approaches.

In this webinar, we demonstrate how to implement these approaches using fable package in R with a real dataset from a hospital.

The registration is open.

Reader (Associate Professor) in Management Science and Business Analytics