Ph.D. in Computer Science at
University of Louisiana at Lafayette,
Center for Advanced Computer Studies (CACS),
M.Sc. in Computer Science at CACS, 2000.
B.Sc. in Electrical Engineering at İstanbul Teknik Üniversitesi, Turkey, 1998.
Dept. of Biology, Room 2168
1510 Clifton Rd.
Atlanta, GA 30322
Research InterestsMy aim is to contribute to the field of computational neuroscience using my existing knowledge in the computer science field. Currently, I am interested in analyzing computer models of single neurons and networks of neurons in the basal ganglia.
One of the challenges posed in our work is to find parameter sets that give biologically realistic results for these models. The difficulty in this task comes from the high dimensionality of the parameter spaces that we use for biologically realistic modeling of neurons. Such neurons are composed of hundreds of compartments, each of which have tens of channels with independent parameters.
Finding adequate parameter settings for these models is not tractable without constraints. An experienced neuroscientist can guide the parameter search by interpreting results and using heuristics. We minimize the search space by clamping groups of compartments and making some channels proportional. One future goal is to develop a experience-based heuristic model of adjusting model parameters to minimize the time it takes to find a solution.
One of our current research areas is to use supercomputers for these parameter search tasks. We try both alternatives of using guided search processes and brute-force search processes. We focus on using the brute-force approach to scan the whole parameter range to better understand the behavior of the model. We generate a database from this search and later analyze it.
In the lab, I use various high-level computational frameworks such as Matlab, GENESIS, and XPP. I also make use of low-level languages such as C/C++, and scripting languages, such as PERL and Python.