The paper "Crowding under Diverse Distance Criteria for Niche Formation in Multimodal Optimization" by Fernández Natalia, Alfonso Hugo, and Gallard Raúl explores the use of niche formation in evolutionary algorithms to address the issue of convergence to a single optimal solution in multimodal function optimization. The authors introduce a crowding method to maintain genetic diversity and prevent genetic drift, which can lead to the population converging to only one of the optimal points. The crowding method involves replacing similar individuals with new ones, ensuring that the population remains spread across multiple peaks in the search space. The study uses two multimodal functions, \( f_1(x) \) and \( f_2(x) \), to evaluate the effectiveness of the crowding method under three distance criteria: phenotypic, genotypic, and fitness. The experiments are conducted with a population size of 200, binary chromosomes of length 30, elitism, and proportional selection. The crowding factors are set to 2, 3, and 4. Key performance metrics include the error of the best individual (Ebest), the error of the population mean fitness (Epop), the number of individuals per niche (Niche Count), and the optimal hits per niche ratio. The results show that the phenotypic distance criterion generally provides the best performance in terms of Ebest and optimal hits, while the genotypic criterion offers higher diversity in the population. The fitness criterion, however, shows higher Epop values but lower diversity. The authors conclude that the phenotypic approach is the most convenient for both types of multimodal functions, offering higher optimal hits and lower Ebest values. Further research will focus on variants of the crowding method and parameter adjustments to enhance diversity.The paper "Crowding under Diverse Distance Criteria for Niche Formation in Multimodal Optimization" by Fernández Natalia, Alfonso Hugo, and Gallard Raúl explores the use of niche formation in evolutionary algorithms to address the issue of convergence to a single optimal solution in multimodal function optimization. The authors introduce a crowding method to maintain genetic diversity and prevent genetic drift, which can lead to the population converging to only one of the optimal points. The crowding method involves replacing similar individuals with new ones, ensuring that the population remains spread across multiple peaks in the search space. The study uses two multimodal functions, \( f_1(x) \) and \( f_2(x) \), to evaluate the effectiveness of the crowding method under three distance criteria: phenotypic, genotypic, and fitness. The experiments are conducted with a population size of 200, binary chromosomes of length 30, elitism, and proportional selection. The crowding factors are set to 2, 3, and 4. Key performance metrics include the error of the best individual (Ebest), the error of the population mean fitness (Epop), the number of individuals per niche (Niche Count), and the optimal hits per niche ratio. The results show that the phenotypic distance criterion generally provides the best performance in terms of Ebest and optimal hits, while the genotypic criterion offers higher diversity in the population. The fitness criterion, however, shows higher Epop values but lower diversity. The authors conclude that the phenotypic approach is the most convenient for both types of multimodal functions, offering higher optimal hits and lower Ebest values. Further research will focus on variants of the crowding method and parameter adjustments to enhance diversity.