Time–Cost Trade-Off Problems with Multi-objective Quasi-Oppositional Teaching Learning-Based Optimization

This paper proposes an improved pattern of the oppositional teaching learning-based optimization (OTLBO). OTLBO utilizes opposition-based learning, a novel machine learning concept in initial population and opposite number generation to have better convergence speed. Instead of opposite numbers, in the present study, quasi-opposite points are utilized and are called quasi-oppositional TLBO (QOTLBO). The proposed approach is performed to acquire Pareto-front (PF) solutions for a well-known 18-activity benchmark problem. When the obtained convergence speed and quality of solutions are considered, this approach appears to be comparatively better, preferable, and deeply satisfying for unraveling time–cost trade-off problems (TCTP) in contrast with the basic NDS-TLBO. Thereby, it can be stated that the developed QOTLBO-based multi-objective approach provides convincing solutions for TCTP optimization problems in construction engineering and management. Moreover, besides this proposed algorithm (QOTLBO), another opposition learning schemes (e.g., comprehensive learning) may also be crucial approach, and this extension is going to be considered in the future research.