My research interests
My Ph.D. research involved building highly realistic computer simulations of
the brain, specifically the rat olfactory (piriform) cortex, tuning the
models to accurately match experimental data, and using these models to
discover interesting ideas about how this system may work. The model I have
built is, to my knowledge, the most accurate computer model of this system
ever constructed. My thesis work uses the model to demonstrate that the
prevailing view on the organization of connections to the piriform cortex and
within the piriform cortex is probably wrong, and proposes an alternative
model which both fits the existing data better and has significant
consequences for the way the system computes.
My other research interests include:
-
Building tools for realistically simulating neurons and networks of
neurons.
I'm a core developer of the GENESIS neural
simulator. My contributions to GENESIS include the synaptic channel code
(the "newconn" library) and the parameter searching library. I have also
written a large number of libraries for simulating the piriform cortex as
part of my thesis research.
- Developing methods to parameterize
complex realistic neuron models.
The methods I have developed so far incorporate a number of optimization
methods, including genetic algorithms, simulated annealing, and
conjugate-gradient descent.
Here is a list of some of my
publications.
More recently, my research interests have shifted away from computational
neuroscience and towards computer science, specifically computer
languages.
|
Go back
to my home page.
|
Last updated January 23, 2008 |
Mike Vanier (mvanier@cs.caltech.edu)