I am interested in algorithms and mathematics for understanding dynamical systems in physics and biology. In particular, I am investigating computational methods for data mining and hypothesizing about the intrinsic physical and chemical relationships that govern many natural systems.

I am part of the Cornell Computational Synthesis Lab (CCSL) at Cornell and work with Hod Lipson. Much of my research includes symbolic regression and related evolutionary algorithms.

I am funded by the National Science Foundation (NSF) Graduate Research Fellowship Program (GRFP), and formerly by the IGERT fellowship program.

I recently had a paper published in Science. Please see the Press Coverage page and the Research page for more information.

Email: mds47@cornell.edu


Eureqa

Try out our new Eureqa automated scientist and let it find hidden formula in your data.



Research

Can computers be made to think like physicists, mathematicians, and engineers? Rather than performing traditional calculations, could they be used for designing models, detecting relationships, and searching for explanations of physical phenomena?
The above podcasts discuss automating science in relation to my work; however, automatic science may not be fruitful if humans aren't able to interpret and understand its results. Instead, I am interested in tools that allow scientists to test their ideas and frameworks more rapidly, allowing computers to determine if these ideas produce accurate and parsimonious models, laws, and predictions automatically.


Press

Some of my research on accelerating science has been covered by popular media outlets. Here are a few selected stories:


Publications

Schmidt M., Lipson H. (2009) "Distilling Free-Form Natural Laws from Experimental Data," Science, Vol. 324, no. 5923, pp. 81 - 85. (see supplemental materials)

Schmidt, M. D., Lipson, H., (2008) "Coevolution of Fitness Predictors," IEEE Transactions on Evolutionary Computation, Vol.12, No.6, pp.736-749.

Schmidt M., Lipson H. (2009), "Symbolic Regression of Implicit Equations," Genetic Programming Theory and Practice, vol. 6, in press.

Schmidt M., Lipson H. (2007), "Learning Noise", Genetic and Evolutionary Computation Conference (GECCO'07), pp. 1680-1685.

(more)