Date: Jan 26th, Thursday
Time: 3:00 pm
Place: Little 339 (the Atrium)
Speaker: Aaron Smith

Title:
Testing for Anti-correlation Using the Binomial Distribution
Abstract:
At times in neuroscience, researchers hypothesis that two structures of
interest
influence each other's behavior even though the their observed behavior is
uncorrelated. One possible hypothesis is that some mechanism prevents
correlation. If this is the case, then neither positive or negative
correlation may  occur over a significant duration. Implying that while one structure's
behavior
slope is constant, the the other structure's behavior slope must change
frequently to prevent correlation. The test uses the binomial distribution to
slope is constant, the the other structure's behavior slope must change
frequently to prevent correlation. The test uses the binomial distribution to
evaluate the alternative hypothesis that there is an anti-correlation by
comparing paired slope changes to random chance.

The pizza and drinks will be provided after the talk.

Pengwen Chen
VP Gator SIAM