Selected Projects
MECHEM - elucidation of reaction mechanisms in chemistry
Ever since about 1904, an important, universal task of chemists has
been to elucidate the pathways or mechanisms of chemical reactions.
MECHEM
is a significant automation of this task. We are now taking steps to
deploy the program on problems from catalytic chemistry, which
presents complicated problems of great industrial importance.
PAULI - discovery of phenomenological selection rules in physics
Physicists invented some quantum properties as a means to explain why
certain reactions among particles occur and others don't. PAULI
automates this task, and has reproduced the discovery of strangeness
using historical data. We have also conjectured and proved a
surprising theorem about this inferential task that concerns the
simplicity of possible solutions.
PENCHANT - discovery of patterned behavior in biology (and other sciences)
Following up on "field work" in cell and development biology, which
uncovered novel patterns in developing Drosophila embryos, we invented
a systematic method for detecting complex patterned behavior in
time-lapse images, or other time-oriented data. We have also applied
this method to data on memory in chess players, and have found new and
interesting patterns. Ongoing applications include nucleations in
crystal growth, photons impinging on a detector in astrophysics, and
eye movements in psychology.
Other newer projects include the elucidation of endocytic pathways in
cell biology, and the discovery of constituent (e.g., quark) structure
in physics.