Thursday, September 09, 2004
Hi. Will we cover any Bayes Net stuff? I'm just curious -- I wouldn't necessarily want to chuck something else in favor of BN. Also, are the course contents more reflective of "foundations of data mining" or "current methods" or "future of data mining"? Do you have a sense of what will be popular in the next 10 years? Have Bayes Nets and SVM and ICA already peaked? Thanks. -- D. Fass
Yes, I will talk about Bayes nets although not in great detail. Concerning your second question, I think its more of a "current methods" class than the other two. Concerning current *research*, I think Bayes nets, SVMs, and ICA continue to attract a lot of attention. As far as which techniques are likely to be in routine use in 10 or 20 years, that's a much trickier question. Clearly the problems these techniques focus on will not disappear but maybe alternative approaches will come to fore. D.
Yes, I will talk about Bayes nets although not in great detail. Concerning your second question, I think its more of a "current methods" class than the other two. Concerning current *research*, I think Bayes nets, SVMs, and ICA continue to attract a lot of attention. As far as which techniques are likely to be in routine use in 10 or 20 years, that's a much trickier question. Clearly the problems these techniques focus on will not disappear but maybe alternative approaches will come to fore. D.
