Researchers have developed an enhanced optimization method termed the Novelty Search enhanced Tree Seed Algorithm to address complex, high-dimensional computational problems. The new algorithm builds upon the foundational Tree Seed Algorithm, an established metaheuristic technique that has demonstrated effectiveness in solving simpler and low-dimensional optimization tasks. According to experimental evaluations, the modified algorithm achieves competitive performance across diverse test suites and delivers superior outcomes in many evaluated cases when applied to high-dimensional challenges.
The standard Tree Seed Algorithm models optimization behavior through interactions between trees and seeds, but its conventional architecture possesses inherent operational constraints. In particular, the traditional method relies on a single evolutionary mechanism between trees and seeds. The authors report that this single evolutionary relationship can restrict both population diversity and population novelty when the algorithm encounters complex, high-dimensional problem spaces, potentially reducing search effectiveness.
To overcome these limitations, the Novelty Search enhanced Tree Seed Algorithm introduces three distinct computational mechanisms designed to enhance search behavior. First, the method incorporates a novelty search framework based on crowding distance to distinguish between sparse regions and crowded regions within the decision space. Second, the algorithm implements a novelty-guided tree migration strategy aimed at reactivating ineffective individuals during the search process. Third, an adaptive step-size seed generation strategy is introduced to improve overall search flexibility during optimization runs.
The research team evaluated the performance of the algorithm across multiple standardized benchmark suites, testing it against functions from the IEEE CEC 2017, IEEE CEC 2020, and IEEE CEC 2022 competitions. The evaluation compared the new method against representative variants of the Tree Seed Algorithm, classical optimization algorithms, and recent metaheuristic approaches. To assess practical applicability, the authors also examined four engineering design problems. Statistical analyses conducted using Wilcoxon signed-rank tests confirmed that the algorithm achieved competitive and frequently superior performance on the tested high-dimensional complex optimization problems.
Subject of Research: Metaheuristic optimization algorithms for complex high-dimensional problems
Article Title: NSTSA: a novelty search enhanced tree seed algorithm with migration and adaptive strategies for complex optimization problems
Article References: Sun, T., Liu, J., Wu, J., Yang, X., Sun, J., Jiang, J., & Li, W. (2027). NSTSA: a novelty search enhanced tree seed algorithm with migration and adaptive strategies for complex optimization problems. Cluster Computing, 30(1), Article 44. https://doi.org/10.1007/s10586-026-06770-x
Image Credits: AI Generated
DOI: 10.1007/s10586-026-06770-x
Keywords: Tree Seed Algorithm, NSTSA, novelty search, metaheuristics, high-dimensional optimization, crowding distance, tree migration strategy, adaptive step-size, seed generation, IEEE CEC 2017, IEEE CEC 2020, IEEE CEC 2022, engineering design problems, Wilcoxon signed-rank test, population diversity, decision space
News Source: Alan Morgan. (November 21, 2026). Novelty Search Tree Seed Algorithm Improves Complex High-Dimensional Optimization Performance. Scienmag.



