USING VARIANTS OF GENETIC ALGORITHM AND LEARNABLE EVOLUTION MODEL TO SOLVE RESOURCE-CONSTRAINED PROJECT SCHEDULING PROBLEM

The resource-constrained project scheduling problem (RCPSP) is a complex scheduling challenge as it is proven to be NP-hard. The Learnable Evolution Model (LEM) is a non-Darwinian evolutionary approach that speeds up convergence using machine learning instead of crossover. It classifies individuals into high-performance (H-group) and low-performance (L-group) based on fitness, learns distinguishing features, and generates new individuals through an instantiation step. To ensure diversity, LEM applies mutation as a Darwinian component, making it more efficient than traditional evolutionary methods. This paper proposes a new approach, which attempts variants of genetic algorithms and LEM, aiming to tackle issues in generating Gantt charts for big cases.