Enter a search term into the box below to search for all events matching those terms.
Start typing a search term to generate results.
How do I use this calendar?
You can click on an event to display further information about it.
The toolbar above the calendar has buttons to view different events. Use the left and
right arrow icons to view events in the past and future. The button inbetween returns you
to today's view. The button to the right of this
shows a mini-calendar to let you quickly jump to any date.
The dropdown box on the right allows you to see a different view of the calendar, such
as an agenda or a termly view.
If this calendar has tags, you can use the labelled checkboxes at the top of the page to select
just the tags you wish to view, and then click "Show selected". The calendar will be redisplayed
with just the events related to these tags, making it easier to find what you're looking for.
Abstract: Trajectory data mining has been of increasing research
interest due to an abundance of modern tracking devices and a number
of real-world applications. Likewise, large-scale network mining has
attracted significant attention as it helps to reveal underlying
dynamics of networks. In our research, we are interested in mining and
analysis of large-scale trajectory networks - networks where the nodes
are moving objects (cars, pedestrians, etc.) and the edges represent
contacts between objects as defined by a proximity threshold. In the
first part of the talk, I will discuss the problem of mining group
patterns in trajectories of pedestrians coming from motion video
analysis. We are interested in interactive analysis and exploration of
group dynamics, including various semantics of group gathering and
dispersion. In the second part of the talk, I will discuss methods for
evaluating the importance of moving objects (nodes) in a trajectory
network. To address these questions we devise fast and accurate
methods for concurrent evaluation of dynamic network metrics including
node degree, triangle membership and connected components for each
moving object, over time. If time allows, I will also discuss our
recent work on network representation learning. Lately, these methods
have been shown to perform very well on static networks. However, in
real-world, networks are continuously evolving. I will discuss a
random-walk based method for effectively learning representations of