Algorithms for the carpool problem

We investigate the carpool problem, on choosing a driver from a subset of a set of people who are members of a carpool. Coppersmith el al. [4] examine the bounds on the values that characterize the fairness of carpool algorithms. In this report, we study the methods introduced in that paper and suggest computational techniques for further improving the lower bound. Specifically, we introduce an algorithm to search the tree of evolution of a given error vector. With this algorithm, we eliminate more possibilities of error vectors from consideration for a higher lower bound.