|
|
| Name |
Bickel, Peter J. |
| Location
|
University of California, Berkeley |
| Primary Field
|
Applied Mathematical Sciences |
| Secondary Field
|
Biophysics and Computational Biology |
 Election Citation
|
|
Bickel is a leading mathematical statistician, making significant contributions to nonparametric inference, sequential analysis, robust methodology, adaptive estimation, and asymptotic theory. He has deep insight into important statistical problems and develops original and powerful methods to solve these problems. |
|
Research Interests
|
|
My current research focuses on both theory and areas of application. My main theoretical concern is understanding why prediction using very high dimensional predictors is as successful as it is despite generally unfavorable theoretical support. This is sometimes cast as the small n large p problem where p is the dimension of data or model and n is the sample size. I have become interested in applications to numerical weather prediction and some questions in astronomy where these issues are very clear. I am also working on various questions in genomics primarily for their scientific interest. The issues I have mentioned certainly arise in this context but analytical understanding seems remote. |
|
|
|