Machine Learning Meetings and Events
Group Meetings: Group meetings are held Mondays from 11am to Noon (talk starts 11:10am) in D.L. Pratt 290C unless otherwise noted. Meetings are coordinated by Hugo Larochelle.
Tea Talks: Tea talks are held every Wednesday at 4:00pm in D.L. Pratt 290C. Talks should be simple, accessible, and not exceed 15 minutes. Speakers bring snacks, make tea, and provide a copy of the presented paper.
Tea Talk Sep 16, 2009: Healing the Relevance Vector Machine through Augmentation, by Carl Rasmussen and Joaquin Quinonero-Candela
- Speaker: Jasper Snoek
- Description: The paper I will be doing for the tea talk on Wednesday will be: 'Healing the Relevance Vector Machine through Augmentation' by Carl Rasmussen and Joaquin Quinonero-Candela from ICML 2005. I have been looking into sparse Bayesian regression recently and I found this paper particularly interesting. This paper is primarily a discussion about why the authors believe the popular Relevance Vector Machine is broken. They propose an interesting fix but admit that in practice it cancels out the computational benefits of the RVM - and suggest that you're better off using a Gaussian process anyway :-)
- Abstract: The paper I will be doing for the tea talk on Wednesday will be: 'Healing the Relevance Vector Machine through Augmentation' by Carl Rasmussen and Joaquin Quinonero-Candela from ICML 2005. I have been looking into sparse Bayesian regression recently and I found this paper particularly interesting. This paper is primarily a discussion about why the authors believe the popular Relevance Vector Machine is broken. They propose an interesting fix but admit that in practice it cancels out the computational benefits of the RVM - and suggest that you're better off using a Gaussian process anyway :-)