HEURISTIC PERTURBATION OF OPTIMIZATION RESULTS IN A DSS FOR INSTRUCTOR SCHEDULING

In this paper we have described a DSS that uses a unique approach of i ntegrating the strengths of management science and human judgement to provide support for a scheduling problem that needs inputs from both t hese areas of human expertise. The manager has the option of accepting a solution developed using optimization technique or intervening dire ctly or indirectly to improve the solution. In intervening directly th e manager makes specific changes to a schedule presented in natural fo rmat. In intervening indirectly the manager specifies the direction of change along with a step size. Knowledge based heuristics act automat ically to revise the schedule in the direction specified. In addition mechanisms have been provided for the user to adapt existing solution to certain kinds of changes in the environment while minimizing the nu mber of perturbations to the existing solution. The DSS has been imple mented for scheduling instructors to executive development programs co nducted by the University of Tennessee, Knoxville.