The quest for efficient sampling algorithms continues to be a demanding research topic due to their wide spread applications. Here, we present an extension of genetic algorithm (GA) to incorporate improved sampling capacity. We develop a fast-navigating genetic algorithm (FNGA) using associated-memory (AM)-based crossover operation which gives more trials with best chromosomes subpart and helps to navigate faster. To mitigate the increased similarity within population, the twin removal genetic algorithm or TRGA is applied. The optimally diverge chromosomes generated by TRGA can introduce potential subpart to enhance the performance of FNGA further. Thus, we combine FNGA and TRGA and named the combination, kite genetic algorithm (KGA). The proposed FNGA and KGA are empirically tested with benchmark functions and the results are found promising. We further employ KGA in the conformational search for the fragment-free protein tertiary structure prediction. The results of ab initio protein structure modelling show that the sampling performance of KGA is competitive.
No takes yet. Share an insight, caveat, or question.
Hoque et al. (2017) studied this question.