Training model based on genetic algorithm and ant colony algorithm

Résumé fourni par la source

To solve the multi-objective conflict, complex constraints and difficulty in balanced optimization of solutions in personalized training planning for college students, this study proposes an optimization model for college student training based on an improved hybrid genetic-ant colony algorithm. This model combines the global exploration capability of the non-dominated sorting genetic algorithm II with elite strategy and the local refinement characteristics of the ant colony algorithm. A dynamic pheromone update mechanism and a parallel collaborative search structure are introduced to construct a progressive optimization framework of "global approximate solution generation-prior knowledge transformation-multi-objective collaborative refinement". Experiments showed that in terms of optimized performance, the per-person training cost was controlled at 1559.7 yuan, and the average skill attainment reached 0.806. In terms of computational efficiency, the single inference time was 1526.7ms, and the data throughput was 662 matching pairs/second. In terms of resource consumption, the calculation amount was 8.57 GFLOPs, the peak memory occupation was 428.3MB, and the energy consumption was 125.6J. In terms of user evaluation, the overall learning experience satisfaction score was 5.88. The proposed model effectively solves the multi-objective and multi-constraint system optimization problems in personalized training planning for college students through the hybrid architecture and collaborative optimization mechanism, and provides a feasible solution for precisely intelligently managing talent training in colleges and universities.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.