In recent years, metaheuristic algorithms have garnered significant attention for their efficiency in solving complex optimization problems. However, their performance critically depends on maintaining a balance between global exploration and local exploitation; a deficiency in either can result in premature convergence to local optima or low convergence efficiency. To address this challenge, this paper proposes an enhanced ivy algorithm guided by a particle swarm optimization (PSO) mechanism, referred to as IVYPSO. This hybrid approach integrates PSO’s velocity update strategy for global searches with the ivy algorithm’s growth strategy for local exploitation and introduces an ivy-inspired variable to intensify random perturbations. These enhancements collectively improve the algorithm’s ability to escape local optima and enhance the search stability. Furthermore, IVYPSO adaptively selects between local growth and global diffusion strategies based on the fitness difference between the current solution and the global best, thereby improving the solution diversity and convergence accuracy. To assess the effectiveness of IVYPSO, comprehensive experiments were conducted on 26 standard benchmark functions and three real-world engineering optimization problems, with the performance compared against 11 state-of-the-art intelligent optimization algorithms. The results demonstrate that IVYPSO outperformed most competing algorithms on the majority of benchmark functions, exhibiting superior search capability and robustness. In the stability analysis, IVYPSO consistently achieved the global optimum across multiple runs on the three engineering cases with reduced computational time, attaining a 100% success rate (SR), which highlights its strong global optimization ability and excellent repeatability.

Accurately predicting the compressive strength of high-performance concrete (HPC) is critical for ensuring structural integrity and promoting sustainable construction practices. However, HPC exhibits highly complex, nonlinear, and multi-factorial interactions among its constituents (such as cement, aggregates, admixtures, and curing conditions), which pose significant challenges to conventional predictive models. Traditional approaches often fail to adequately capture these intricate relationships, resulting in limited prediction accuracy and poor generalization. Moreover, the high dimensionality and noisy nature of HPC mix data increase the risk of model overfitting and convergence to local optima during optimization. To address these challenges, this study proposes a novel bio-inspired hybrid optimization model, AP-IVYPSO-BP, which is specifically designed to handle the nonlinear and complex nature of HPC strength prediction. The model integrates the ivy algorithm (IVYA) with particle swarm optimization (PSO) and incorporates an adaptive probability strategy based on fitness improvement to dynamically balance global exploration and local exploitation. This design effectively mitigates common issues such as premature convergence, slow convergence speed, and weak robustness in traditional metaheuristic algorithms when applied to complex engineering data. The AP-IVYPSO is employed to optimize the weights and biases of a backpropagation neural network (BPNN), thereby enhancing its predictive accuracy and robustness. The model was trained and validated on a dataset comprising 1030 HPC mix samples. Experimental results show that AP-IVYPSO-BP significantly outperforms traditional BPNN, PSO-BP, GA-BP, and IVY-BP models across multiple evaluation metrics. Specifically, it achieved an R² of 0.9542, MAE of 3.0404, and RMSE of 3.7991 on the test set, demonstrating its high accuracy and reliability. These results confirm the potential of the proposed bio-inspired model in the prediction and optimization of concrete strength, offering practical value in civil engineering and materials design.

This study presents an improved variant of the greater cane rat algorithm (GCRA), called adaptive and global-guided greater cane rat algorithm (AGG-GCRA), which aims to alleviate some key limitations of the original GCRA regarding convergence speed, solution precision, and stability. GCRA simulates the foraging behavior of the greater cane rat during both mating and non-mating seasons, demonstrating intelligent exploration capabilities. However, the original algorithm still faces challenges such as premature convergence and inadequate local exploitation when applied to complex optimization problems. To address these issues, this paper introduces four key improvements to the GCRA: (1) a global optimum guidance term to enhance the convergence directionality; (2) a flexible parameter adjustment system designed to maintain a dynamic balance between exploration and exploitation; (3) a mechanism for retaining top-quality solutions to ensure the preservation of optimal results.; and (4) a local perturbation mechanism to help escape local optima. To comprehensively evaluate the optimization performance of AGG-GCRA, 20 separate experiments were carried out across 26 standard benchmark functions and six real-world engineering optimization problems, with comparisons made against 11 advanced metaheuristic optimization methods. The findings indicate that AGG-GCRA surpasses the competing algorithms in aspects of convergence rate, solution precision, and robustness. In the stability analysis, AGG-GCRA consistently obtained the global optimal solution in multiple runs for five engineering cases, achieving an average rank of first place and a standard deviation close to zero, highlighting its exceptional global search capabilities and excellent repeatability. Statistical tests, including the Friedman ranking and Wilcoxon signed-rank tests, provide additional validation for the effectiveness and importance of the proposed algorithm. In conclusion, AGG-GCRA provides an efficient and stable intelligent optimization tool for solving various optimization problems.

While multi-view unsupervised feature selection (MUFS) has achieved substantial progress, real-world unbalanced and incomplete multi-view data with uneven cross-view missing proportions poses significant challenges. Non-uniform missingness undermines cross-view complementarity and triggers feature selection bias, wherein highly complete views dominate feature evaluation, leading to suboptimal feature subsets, degraded convergence performance, and excessive computational overhead. To tackle this problem, we propose WTUIMUFS, a two stage MUFS method integrating fractal wave-optics optimization (FWOO) and tensor anchor graph joint learning. In the first stage, a Fractal WOO (FWOO) is developed for reliability-aware feature space exploration. By integrating fractal phase-shift perturbation with intensity-guided adaptive updating, FWOO enhances global search capability and alleviates premature convergence under heterogeneous view completeness. In the second stage, a Tensor Anchor Graph Joint Learning model is constructed to simultaneously perform view-aware missing value completion and tensor-guided latent representation modeling, thereby restoring high-order cross-view correlations disrupted by incomplete observations. Additionally, a view-completeness-aware dynamic reweighting mechanism is embedded throughout both stages to adaptively balance view contributions based on observation ratios and reconstruction reliability, mitigating dominance from highly complete views while preventing over-reliance on heavily imputed ones. Extensive experiments on eight benchmark datasets demonstrate that WTUIMUFS consistently outperforms eight state-of-the-art methods in terms of five evaluation metrics, convergence behavior, and runtime efficiency. Under varying missing ratios, average improvements over the second-best method reach 3.85% (ACC), 3.45% (NMI), 4.28% (Purity), 3.21% (F-score), and 2.78% (ARI), confirming the effectiveness, robustness, and computational efficiency of the proposed method.