MLSG: Stanford ML course

[ambarisha b]

Hi,

Stanford has started enrollment for the machine learning course as planned (http://ml-class.org)

What do you think of syncing our group activity with the course schedule? I think its a great idea. If ML were as simple as listening to 10 min lectures, everyone would be a scientist. But its not. A group will really help. So, I think structuring our work along the course schedule is a great idea. We can discuss the lectures and problem sets during the meets. What do you say?

Also, what should be the topic of discussion for our next meet?

//Ambarish

[Pranav Agarwal]

We should discuss on Multiple Kernel Learning and Mathematics behind it. Last time, we did cover Naive Bayes and few small discussions on different areas in CS.

I believe what we should concentrate more on, is actually starting on implementing an idea. More details on this in our next session. As for the ML-Classes, definitely a good idea but we are already ahead of them. KKT and Multiple Kernel Learning comes in 13th Lecture as per their schedule. They are focusing on Regressional Models right now. But, i agree on specifically doing those assignments in groups. It would help a lot.

Next Session, this Friday? Room booked?

Pranav

[ambarisha b]

Today's meet is at 8PM in room no 1232