580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How can novel swarm intelligence algorithms inspired by specific animal behaviors enhance optimization performance in engineering applications?
This research area investigates developing and validating new swarm intelligence algorithms based on the unique social and foraging behaviors of specific animal species to improve accuracy, convergence speed, robustness, and stability in solving global optimization problems, particularly in engineering contexts.
Key finding: Proposed the Sparrow Search Algorithm (SSA) inspired by sparrows' foraging and anti-predation behaviors, demonstrating superior accuracy, convergence speed, stability, and robustness compared to Grey Wolf Optimizer (GWO),... Read more
Key finding: Developed enhanced versions of the Salp Swarm Algorithm (SSA), including population diversification (SSA std) and self-adaptive parameter tuning via genetic algorithm (SSA GA-tuner), resulting in significantly improved... Read more
Key finding: Introduced the Bees Algorithm (BA), inspired by the natural foraging behavior of honey bees, combining exploitative neighborhood search with random exploratory search; demonstrated competitive or superior performance over... Read more
Key finding: Proposed an enhanced Firefly Algorithm incorporating fireflies' gene exchange, pheromone communication, and environmental effects like wind on pheromone dispersion, outperforming the traditional Firefly Algorithm and standard... Read more
Key finding: Developed a hybrid swarm intelligence algorithm (HSI) combining Bat Algorithm (BA) and Practical Swarm Optimization (PSO) to address large-scale global optimization benchmarks from IEEE CEC'17; demonstrated promising... Read more
2. What roles do exploration-exploitation balance strategies and adaptive parameter control play in improving particle swarm optimization (PSO) performance for global optimization?
This research theme examines how different learning strategies, adaptive parameter tuning, hybridization, and neighborhood topologies within PSO variants can effectively address challenges such as premature convergence, local optima trapping, and scalability in optimization problems, enhancing both convergence speed and solution quality.
Key finding: Proposed Dual-Stage Hybrid Learning PSO (DHLPSO), partitioning iterations into two stages emphasizing exploration via Manhattan distance based learning selecting the furthest and better particles, and exploitation through... Read more
Key finding: Reviewed recent advances in SI algorithms PSO, ACO, ABC, and FA for scheduling and optimization in cloud computing; emphasized the importance of parameter setting effects and the growing utilization of hybrid swarm techniques... Read more
Key finding: Introduced a steady state PSO variant inspired by the Bak-Sneppen co-evolution model, updating only the least fit particle and its neighbors asynchronously each step; experimental analyses revealed significant improvements in... Read more
Key finding: Provided a comprehensive critical analysis of SI-based algorithms focusing on their mimicry of evolutionary operators (mutation, crossover, selection), exploration-exploitation mechanisms, and modeling via dynamical systems... Read more
Key finding: Surveyed the theoretical foundations of SI-based metaheuristics, illustrating how the stochastic iterative search processes relying on local interactions and historical memory achieve global optimization via... Read more
3. How can swarm intelligence algorithms be hybridized or integrated with decision-making and classification frameworks to solve complex many-objective optimization and classification problems?
This theme explores the integration of swarm intelligence algorithms with multi-criteria decision analysis (MCDA) methods, ordinal classifiers, and local search strategies to address solution selection challenges in problems with many objectives or high-dimensional data, improving solution quality in target regions and preventing premature convergence in classification and optimization.
Key finding: Proposed augmenting Multi-objective Grey Wolf Optimization and Indicator-based Multi-objective Ant Colony Optimization with ordinal classification-based decision maker preference modeling to bias search toward the Region of... Read more
Key finding: Introduced a novel Simplified Swarm Optimization (SSO) algorithm combined with an Exchange Local Search (ELS) strategy that refines solutions in the neighborhood of current particles; empirical results on 13 widely used... Read more
Key finding: Provided comprehensive coverage of diverse nature-inspired swarm intelligence algorithms and their theoretical underpinnings coupled with practical applications across engineering, economics, and social sciences; emphasized... Read more