Emaneul Parzen
Texas A&M University
Polymath Data Analysis, Simulation, Goodness of Fit, and Condtional
Quantile Estimation.
This paper presents concepts and philosophy of "polymath data
analysis" and statistical methods mining. Our goal is to develop
effective methods to analyze data by models for data that are good
for simulation. Especially novel: Bayes theorem for quantile
functions, application to estimation of conditional quantile
functions and identification conditional quantiles at percentiles
u=.1, .25, .75, .9. We use logistic regression to smooth scatterplots.
Back to Colloquium Series