Compstat 2002, Berlin

Competition for the Best Presented Poster
prize donated by the ASC

Background

After a successful poster competition at Compstat 2000 in Utrecht, the ASC committee decided to repeat the offer of a prize to encourage participants presenting material through posters to give more attention to the design and presentation aspects of their work. The motivation and criteria behind this proposal are discussed on the main Poster Prize page.

Criteria

The posters were judged on the quality of their design and their effectiveness at communicating their message. The nature or quality of the substantive content of the poster was not considered as a factor.

The prize for this competition was €150.

The Poster Competition

There were 35 posters presented at Compstat, in two sessions. The judges were:

Andrew Westlake, representing the ASC

John Hinde, representing the European Regional Section of IASC

Wolfgang Haerdle, representing the Compstat local organisers

The judges viewed the posters independently, and then came together to reach a decision about the prize. Compared with Utrecht there seemed to be a slight overall improvement in quality, perhaps the result of better publicity to the authors of the existence of the prize. More than half of the authors seemed to have made some specific effort to respond to the challenge of the prize, by introducing eye-catching components or by using the two-dimensional space. However, there are still none that reach the highest standards seen at some conference in other disciplines.

The Winner

Only a small number of posters were in contention for the prize, and the judges were unanimous in agreeing to award to prize to:

Isabel Serrano, Juan Jose Hoyo & Dolores Gonzŕlez, University of Huelva, Spain Non linear state space models and Gibbs sampling applied to fisheries stock assessment This poster has an eye-catching design and strong structure, though rather too much material. Click here for a larger image.
Highly Commended The following three posters were highly commended by the judges. None were in final consideration for the overall prize, but all were particularly strong in some respect.
An Algorithm to detect multiple change-points in the residual variance of financial time series Andreas Stadie, University of Göttingen Very clever idea, but implementation could be better
Estimation of dynamic models using kernel density Xavier Font, Manuel Marti & Pilar Muńoz, EUPMT & UPC, Spain A clean and simple design that used the space well
Extracting implicit density functions from short-term interest rate options Hannah Nielsen, Freie University, Berlin Not the strongest design, but executed well