When I'm not an academic...
In my free time, I like to listen to music, play guitar, watch television shows, drink beer and eat food that other folks cook. As far as music is concerned, I like (mostly) guitar-based genres. Some bands I like are Radiohead, Spoon and Belle & Sebastian. I generally learn to play music to which I listen and similar tunes. Lately, I have also been experimenting with hooks, similar to The Apples in Stereo. My favorite television shows are South Park and Futurama. I also like British shows, such as The Office, Spaced and The I.T. Crowd. I enjoy a variety of beers. Delirium Tremens, a yeasty Belgian pale ale, tops my list. Some other beers I enjoy are Southern Pecan, New Castle and Shiner Bock. While I don't cook much myself, several of my friends often experiment in the kitchen and on the grill. A recent innovation is hot dog (preferably Hebrew Nationals), covered in hamburger meat, wrapped in bacon and sprinkled with cheese... seven thousand calories of deliciousness.
I am a Ph.D. Candidate in Computer Science and Engineering at Mississippi State University. My research interests include Bayesian networks, heuristic search, algorithms and computational biology. My main advisor is Changhe Yuan; I have also worked very much with Susan Bridges. My other committee members are Eric Hansen, Zhaohua Peng and Andy Perkins. I also work with Abiola Olaniyan, an REU student in CSE, Zhifa Liu, a CSE master's student, and Feng Tan, Zhaohua's Ph.D. student. I plan to graduate in either May or August 2012.
I am currently lecturing for one section of CSE 1284, our CS1 course. There are about 50 students in the course. I am also taking CSE 8990, an introductory model checking course with Eric Hansen. It is pretty large for a graduate course; there are about 20 students. I was selected to participate in the Preparing Future Faculty program at MSU. Most of my remaining productive time is consumed by research. In addition to my dissertation about learning optimal Bayesian network structures, I am also exploring approximate methods for structure learning, classification structure learning and applications to molecular genetics.
Recently, I completed the Preparing Future Faculty program at Mississippi State. It was a great program, and I highly suggest anyone interested in a career in higher education take this or a similar program. One of the things I found most fascinating was learning about how computer science compares to other disciplines. About half of the program was dedicated to research (establishing a successful lab, writing grants, etc.), and the other half was given to teaching. Despite my love of the Professor in Futurama, I do not want to be "a professor"; I want to learn how to effectively communicate with all students. Consequently, I hope to lecture in some upper level courses while at the University of Helsinki, such as algorithms or artificial intelligence.
I have successfully defended my dissertation entitled "Learning Optimal Bayesian Networks with Heuristic Search" and begin a post-doc at the Finnish Center of Excellence in Computational Inference Research (COIN) in July. My research will include Bayesian network structure learning and computational biology in Petri Myllymaki's group. The position is for three years. While there, I hope to expand into other research areas, including inference and bounded error algorithms. I am particularly interested in critical points of NP-hard optimization problems. Some prelimiary research suggests that structure learning may have a critical point; we know the problem is solvable in polynomial time if each variable has at most one parent (Chow-Liu), and we know that the problem is NP-hard if each variable has at most k parents (Chickering), k >= 2. However, there has been no average case analysis when the number of parents is, for example, 1.5.