Last week, it was our last class with Prof. Foster Provot for this semester. This is a PhD level seminar discussing all kinds of topics related to data mining and machine learning. As the only three registered PhD students in Stern, Xiaohan, Mihaela and I were "pushed" to give (bi-)weekly presentations and lead discussions for every paper on those topics.
Oh, god! That was hard! I couldn't understand this. When I sit down in the class and listen to the professors, they are all talking and smiling, making all kinds of jokes, writing gracefully and drawing nice pictures on the board. They are teaching as if doing something really really really easy. However, when I stood in front of the class, no matter how hard I had prepared, I felt nervous, awkward, and then suddenly forgot what I should say. My tones got wondered and my voice became frozen. My confidence was quickly fading out... In fact, I was pretty confident in my presentation skills because I already had some conference/workshop presentation experiences before. I always felt proud of my cool behavior in front of a group of people. But now the truth was that it did not work here! Teaching in class is totally different from giving a short 20-minute talk, at all! For this, I really admire Foster! He is such a sharp person and a great professor. He can always notice the key point in our thoughts and help us sort it out right away. Often times, his questions are actually helpful and informative "hints", which inspire us to think what we have neglected and then better organize our thoughts.
Prof. Anindya Ghose once told me that when you talk to people, you should try to make your point as clear as you can at the first time. Do not wait for people to find themselves confused and then ask you. I believe this is important, but it is not easy to achieve. Sometimes, when we explain something, we have a tendency to either describe it too much that makes the redundancy, or speak too little that leads to the ambiguity. (It seems that the distribution for the intensity of our explanatory words is "bimodal", either too high or too low.) I like Prof. Panos Ipeirotis's teaching, because his way is highly logic. You feel like you are led into a room, and then get to explore by yourself with encouragements time by time. He does not show the whole picture at one time, but leave to us ourselves to find it out. That is coolest part. You never know how big the picture is! Just like an adventure game!
I sometimes was imaging myself in the future, can I do this well when I become a real professor? Will my students enjoy my teaching too? Yea, I believe so! That is my goal and just keep going:-)
Saturday, May 31, 2008
A taste of teaching
Tuesday, May 20, 2008
Data Mining Blogs: The Big List(ZZ from Sandro Saitta)
Sandro Saitta has a full list about the data mining blogs. Just something very nice that can be introduced here:)
: this blog gives news about data mining and AI very frequently (Alberto Roldan): the blog of the data mining laboratory at Brigham Young University, mainly about social communities and meta-learning (Data Mining Lab): discuss statistics and predictions among other interesting topics (John Aitchison): comprehensive posts on technology and news related to data mining and machine learning. Also a lot of very useful resources (Pete Skomoroch): although not updated recently, this blog has interesting posts about data visualization and statistics (Donald Farmer): a blog that focus on data mining using Microsoft SQL Server (Jamie Mac): data mining with a point of view from statistics (Rachel Graham): a personal view on data mining with posts on different applications and news (Shane Butler): a machine learning blog from a PhD student at Cambridge (Jurgen Van Gael): more machine learning oriented but contains a lot of useful information (Pierre Dangauthier)
