Medal awards 2000
- PDF / 48,259 Bytes
- 2 Pages / 595 x 842 pts (A4) Page_size
- 118 Downloads / 219 Views
#2001 Operational Research Society Ltd. All rights reserved. 0160-5682/01 $15.00 www.palgrave.com/jors
Medal awards 2000 Goodeve and President's medals From 1998 onwards, the eligibility for these medals has been extended to include articles in any of the Society's publications, ie the Journal, the European Journal of Information Systems, OR Insight and the Newsletter. They are awarded for papers published in the previous year. From 2000, at the initiative of President Mike Pidd, the President's medal will now be awarded for the best account of OR practice. Authors whose papers are shortlisted are invited to present their papers at the ORS annual conference. The winner is chosen by a panel of judges, with some input from conference delegates. Thus in 2000 two President's medals have been awarded, one under the previous criteria for a paper published in 1999 and the other for a presentation at the ORS conference in 2000. 2000 Goodeve medal The Goodeve medal for 2000 is awarded to J Banasik, JN Crook and LC Thomas of the University of Edinburgh for their paper: `Not if but when will borrowers default?' (J Opl Res Soc (1999) 50: 1185±1190). Credit scoring systems are based on OR and statistical models and are typically built to answer the question, `How likely is an applicant for credit to default by a given time in the future?' A wide range of techniques can be used to answer this question, but the underlying methodology is to take a sample of previous customers and classify them into `good' and `bad' depending on their repayment performance over a given period. Unfortunately this approach does not take into account the dynamic element to credit statusÐthe fact that borrowers default at different times in their credit history. The question which should be asked is not simply if an applicant will default, but if they default when will this be? Banasik, Crook and Thomas addressed the problem by developing models that could provide answers to this question. Essentially they developed three models all based on Cox's proportional hazards model (PHM). These were a non-parametric model, a model using proportional estimators with an exponential hazard rate (which corresponds to an accelerated life model), and a PHM with Weibull baseline. Although the application of PHM in this context is not entirely original, the authors succeeded in presenting this important application in a way that will enable practitioners to appreciate the usefulness of survival
analysis in credit scoring. The application of the models to real life data and comparison of the models with a standard credit scoring model based on logistic regression, demonstrates the value of the proposed approach. Above all this paper illustrates how one can utilise data more ef®ciently in addressing credit scoring. For all the above reasons and for the belief that this work can contribute signi®cantly to the advancement of the application of OR and statistical techniques in credit scoring, the authors are awarded the 2000 Goodeve medal. 2000 President's medal (old crit
Data Loading...
