Henry Kautz: Understanding Human Behaviour from Sensor Data
The convergence of advances in algorithms for probabilistic reasoning and the development of low-cost, easily-deployed sensors is reviving the dream of AI to develop systems that can understand the narrative of ordinary human life. On the reasoning side, the AI community is developing techniques that bridge the gap between propositional Bayesian representations and hierarchical models of goals, plans, and actions. On the sensing side, new technologies such as RFID tags, GPS, motes, and wearable multi-modal sensors allow us to gather direct information about many aspects of human experience. I will describe recent work with my students and colleagues on developing systems that learn patterns of human activity for everyday tasks, both indoors and outdoors, using a variety of dynamic probabilistic models. I will then describe applications of these techniques to healthcare systems as part of the Assisted Cognition Project, a joint effort between our departments of computer science and rehabilitation medicine.
More information: http://www.cs.ubc.ca/events/seminars/csicics.shtml
When:
Thursday, March 16, 2006 - 15:00 to 16:30
Where:
DMP 310
Tags:
When: to
Where: DMP 310 - 6245 Agronomy Rd, Vancouver, BC, V6T 1Z4
Traditionally communications network has been considered synonymous with connectivity. Focus for network research and development was majorly on bandwidth, availability and protocol interoperability. While these factors continue to play an important role in the future network R&D, a new set of business challenges and opportunities such as mission critical business applications, self-defending security networks, and network as a distributed computing platform, is emerging and requires the attention of network researchers and developers.
For those of you who didn’t attend this extremely popular event last year, In My Experience is basically the straight-dope on Computer Science. Hear from several CS students, in all stages of their degree and what they are doing. If you have that burning question, like, “Should I go into co-op?” (of course!), or “Should I do a double major or honours?” (well, duhhh…), this is the place to get your answers - from peers who’ve had to make those same choices.
FSMLabs produces RTLinux, the popular real-time operating system. Dr. Yodaiken is in Vancouver specifically to announce the formal opening of the company’s newly-established Vancouver office.
It is often preferable to build a storage service out of a collection of individual components rather than out of a single monolithic server. The reasons for preferring such a structure are increased fault-tolerance, scalability, and economics. However, the distributed nature of the system can make it complex to design, difficult to implement, and nearly impossible to test making the whole enterprise a daunting one. One of the ways of reducing the complexity of distributed storage systems is to first design a set of simple abstractions and services that can then be composed easily to provide the required functionality. The talk will illustrate examples of this technique with reference to some systems – Petal ‘96, Frangipani ‘97, Boxwood ‘04, and Eclipse ‘05– on which I have worked. These systems demonstrate that it is feasible to build complex storage systems using a set of building blocks that are easy to design, implement, and test.