Tighter Bounds on Non-clairvoyant Parallel Machine Scheduling with Prediction to Minimize Makespan

Abstract

This paper investigates the non-clairvoyant parallel machine scheduling problem with prediction, with the objective of minimizing the makespan. Improved lower bounds for the problem and competitive ratios of online algorithms with respect to the prediction error are presented for both the non-preemptive and preemptive cases on m identical machines.