Analysing sub-daily time series data

Rough agenda:

• 5:45: Pizzas, drinks & networking

• 6:30: Intro & main presentation

• 7:30-8: More networking

Time series analysis used to involve mostly annual, quarterly and monthly data. Now it increasingly involves daily, hourly, or even more frequent observations. However, the available tools have not kept pace with the needs of analysts, and often we end up trying to use inappropriate tools to visualize, forecast and analyse time series.

In response to this need, several people in the Monash research group led by me and Di Cook have been working on some new tools to help with analysing modern high-frequency time series. We will discuss some new R packages for managing time series objects, visualizing time series and forecasting time series. Our focus is on hourly and half-hourly data, although the tools are general enough to handle a much wider range of frequencies.

We will introduce tsibbles (tibbles for time series), demonstrate some cool new calendar-based graphics, and show off our new FASSTER forecasting model.

Rob J Hyndman is Professor of Statistics in the Department of Econometrics and Business Statistics at Monash University. He is also Editor-in-Chief of the International Journal of Forecasting and a Director of the International Institute of Forecasters. Rob is the author of over 150 research papers and 5 books in statistical science. In 2007, he received the Moran medal from the Australian Academy of Science for his contributions to statistical research, especially in the area of statistical forecasting. For over 30 years, Rob has maintained an active consulting practice, assisting hundreds of companies and organizations around the world. He has won awards for his research, teaching, consulting and graduate supervision.

Rob will be joined by two of his students:

Earo Wang: Earo is currently doing research on statistical visualisation of temporal-context data, as part of her PhD at Monash University. She enjoys developing open-source tools with R, and is the (co)author of some widely-used R packages including anomalous, hts, sugrrants, rwalkr and tsibble. Earlier this year, she was described by Yihui Xie (author of rmarkdown, knitr, bookdown and blogdown) as "one of the most impressive R ladies I have ever met".

Mitchell O'Hara-Wild: Mitchell O'Hara-Wild is an honours student studying econometrics at Monash University. His research primarily involves time series forecasting, which includes the development of the new FASSTER model for handling multiple seasonalities. Mitchell teaches data analysis using R at Monash University, has authored multiple R packages and assists with the development of the forecast package.