MTTA: Modernized Tiki-Taka Algorithm with Role Specialization for Solving Engineering Application Problems and Feature Selection

With the growing complexity of engineering optimization and high-dimensional data analysis tasks, balancing global exploration and local exploitation remains a core challenge in computational intelligence. The Tiki-taka Algorithm (TTA), a football-inspired metaheuristic, is simple to implement and provides competitive baseline performance. However, it may suffer from premature convergence, rapid loss of population diversity, and rigid search transitions when solving complex multimodal or high-dimensional problems. This paper proposes the Modernized Tiki-taka Algorithm (MTTA), which incorporates a role-specialization mechanism inspired by contemporary football tactics into the optimization framework. MTTA adopts a Logistic–Tent hybrid chaotic mapping for uniform initial population distribution, and establishes a fitness-based three-role mechanism (forwards, midfielders, defenders) with tailored update rules, achieving a smooth, adaptive balance between exploration and exploitation throughout the iteration to fundamentally overcome TTA’s inherent flaws. Comprehensive experiments on classical benchmarks, the IEEE CEC 2017 test suite, constrained engineering problems, and feature selection tasks demonstrate that MTTA achieves statistically significant superiority over TTA, classic metaheuristics, and state-of-the-art optimizers in convergence speed, solution accuracy, and robustness, providing an efficient, scalable solution for complex real-world optimization scenarios.