| 摘要 |
A fair incentive calculating method with a reasonable amount of computation is performed by inputting a set of x jobs X={1, 2, . . . , x} and processing times of each worker {t(i)} (1≦̸i≦̸n) and providing a set of jobs W(A, X, t(i), t(−i)) assigned to each worker i, where A is a given scheduling algorithm; determining MaxInc(X, t(i)) by recursively solving MaxInc(X, t(i))=&Sgr;_{w&egr;X} t(i, Kind(w))+max_{s(i) &egr;Slower(t(i))} MaxInc(X, s(i)); and using the step of determining MaxInc(X, t(i)) to calculate an incentive NewInc(A, X, t(i), t(−i)) for each worker i based on NewInc(A, X, t(i), t(−i))=&Sgr;_{w&egr;W(A, X, t(i), t(−i))} t(i, Kind(w))+max_{s(i) &egr;Slower(t(i))} MaxInc(X, s(i)). |
