Rob Hyndman: Visualization of big time series data
Details
---------------------------------------------------------------
Timing for the evening
5pm - we have access to the Arena if you want to arrive early and reserve your pew. There is a cafe just outside.
6pm - we will start
7:45 - join us upstairs at platform28 for drinks - we have a tab behind the bar (only for the cheap beers!)
Note: The reputation of the Statisticians precedes them, and it has been deemed necessary to have a security guard to keep them under control. Say Hi to her on the way out of the Arena!
----------------------------------------------------------------
This is a joint event between Data Science Melbourne and our friends from SSA Vic (https://www.meetup.com/Statistical-Society-of-Australia-Victorian-Branch/).
Our guest for the evening is Rob Hyndman (http://robjhyndman.com/), an Australian statistician known for his work in forecasting.
After the talk please join us for drinks at Platform 28 (http://platform28.com.au/), which is just around the corner.
For rough times, see further down this page - we will commence at 6pm and be getting the last train home!
------------------------------------------------------------------
Visualization of big time series data
It is becoming increasingly common for organizations to collect very large amounts of data over time. Data visualization is essential for exploring and understanding structures and patterns, and to identify unusual observations. However, the sheer quantity of data available challenges current time series visualisation methods.
For example, Yahoo has banks of mail servers that are monitored over time. Many measurements on server performance are collected every hour for each of thousands of servers. We wish to identify servers that are behaving unusually.
Alternatively, we may have thousands of time series we wish to forecast, and we want to be able to identify the types of time series that are easy to forecast and those that are inherently challenging.
I will demonstrate an approach to this problem using a vector of features on each time series, measuring characteristics of the series. For example, the features may include lag correlation, strength of seasonality, spectral entropy, etc. Then we use a principal component decomposition on the features, and plot the first few principal components. This enables us to explore a lower dimensional space and discover interesting structure and unusual observations
About the speaker: Rob J Hyndman is Professor of Statistics in the Department of Econometrics and Business Statistics (http://www.buseco.monash.edu.au/depts/ebs/) atMonash University (http://www.monash.edu.au/) and Director of the Monash University Business & Economic Forecasting Unit. He is also Editor-in-Chief of the International Journal of Forecasting (http://www.forecasters.org/ijf/) and a Director of the International Institute of Forecasters (http://www.forecasters.org/). Rob is the author of over 100 research papers 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 25 years, Rob has maintained an active consulting practice, assisting hundreds of companies and organisations. His recent consulting work has involved forecasting electricity demand, tourism demand, the Australian government health budget and case volume at a US call centre.
http://photos4.meetupstatic.com/photos/event/d/1/e/e/600_437213742.jpeg
------------------------------------------------------------------
The Stadium:
https://www.youtube.com/watch?v=8rXO7CMn7nQ
Note the venue has 200 seats, so its first in best dressed, although plenty of standing room in the Chaddy End
https://www.youtube.com/watch?v=jgDIHnHTars
The Warm Up: 5:30pm - doors open, come and mingle
The Match: 6:00pm - talks commence
https://www.youtube.com/watch?v=cbu_KLZlCIU
Transfer Window: Jobs on offer - to announce a job it must be posted in the discussions BEFORE the meetup
https://www.youtube.com/watch?v=acoIUUxTXmA
The Final Whistle: 8:15pm @ latest - we must vacate the NAB
https://www.youtube.com/watch?v=sHO_Jo-ma3s
The Post Match Assessment: 8:00pm ++ resume at Platform 28 for some deep discussions over drinks and dinner with Rob
https://www.youtube.com/watch?v=3Es-RIBnba8
Getting to the Venue
The easiest way is to go up the Southern Cross Station stairs (Pic 2), which is located at the corner of Bourke St & Spencer St. Once you are at the top, walk across the bridge towards Etihad Stadium and the first building on your right is the NAB building (Pic 3). When you've entered the building the Arena is just on your left.
Or more instructions are here...
http://www.nabvillage.com.au/directions
http://photos3.meetupstatic.com/photos/event/a/9/2/a/600_436903306.jpeg
http://photos2.meetupstatic.com/photos/event/8/d/7/a/600_435876218.jpeg
Stairs at Spencer St
http://photos1.meetupstatic.com/photos/event/8/d/7/d/600_435876221.jpeg
The NAB is just at the other end of the footbridge
http://photos3.meetupstatic.com/photos/event/2/a/2/c/600_437410796.jpeg
Go through the revolving doors
http://photos3.meetupstatic.com/photos/event/2/a/3/3/600_437410803.jpeg
The entrance to 'The Arena' is then just on the left
http://photos3.meetupstatic.com/photos/event/8/e/4/a/600_435876426.jpeg
250 chairs but lots of room.
The Post Match Assessment
Please join us for complimentary (upto our tab!) beer/wine and soft drinks at Platform 28 after the talk. On exit from NAB, there is a shortcut under the Medibank building that takes you right there.
http://photos2.meetupstatic.com/photos/event/a/7/b/f/600_436902943.jpeg
http://photos4.meetupstatic.com/photos/event/2/a/5/1/600_437410833.jpeg
Take the above shortcut which is right opposite the exit to the NAB
http://photos2.meetupstatic.com/photos/event/2/a/8/1/600_437410881.jpeg
At the bottom you can see Platform 28 just across the street
http://photos4.meetupstatic.com/photos/event/2/a/9/e/600_437410910.jpeg
Looks good eh? We have the upstairs room booked
Rob Hyndman: Visualization of big time series data