ps I dont see the paper on your http://cseweb.ucsd.edu/%7Eelkan/#papers could I post it on my wiki for my grad students or is it posted elsewhere or still undercover?
>Doug, I look forward to talking! My most recent paper in this area is
attached. Comments very welcome! I will be in Dubai and India Dec. 10 to 23, but maybe we can do lunch the week after Xmas, since you are local?
If you have an algorithmic talk you would like to give, there is one slot open in the seminar series I am organizing next quarter: http://ucsdai.blogspot.com/
Best, Charles 2012 talk at UCSD
Link prediction via matrix factorization. Aditya Krishna Menon, Charles Elkan. In Machine Learning and Knowledge Discovery In Databases - European Conference, ECML-PKDD, Proceedings Part II, 2011. [pdf]
Fast algorithms for approximating the singular value decomposition. Aditya Krishna Menon, Charles Elkan. To appear in Transactions of Knowledge and Data Discovery: Special Issue on Large-Scale Data Mining (TKDD-LDMTA), 2010. [pdf] code
A log-linear model with latent features for dyadic prediction. Aditya Krishna Menon, Charles Elkan. In IEEE International Conference on Data Mining (ICDM), Sydney, Australia, 2010. [pdf] [slides] [code]
- AI Seminar, UCSD 2012 3/5/12: Open Causal analysis from survey data with network autocorrelation. ?? possible title