An Improved Genetic Algorithm for Web Phishing Detection Feature Selection
Jiachen Wang · 2022 Asia Conference on Algorithms, Computing and Machine Learning (CACML) · 2022
Feature selection is a useful dimension reduction method in data preprocessing for machine learning. It is an discrete optimization problem in essence. Genetic algorithm is a meta heuristic optimization algorithm for discrete optimization problem, which simulates the biological evolution behavior in nature. The original Genetic algorithm, like many other swarm intelligence optimization algorithms, has its own defects, easy to fall into local optimum, and slow convergence speed. To improve the performance of Genetic algorithm, an improved algorithm ECGA is proposed. In order to measure the diversity of population, the information entropy of population is calculated after fitness calculation. To speed up convergence, the consensus mechanism is presented after mutation which is used as a alternate mutation operator. And finally ECGA is applied to feature selection to test its actual effect. The experimental results show that the improved Genetic algorithm is better than the feature selection algorithm proposed in recent 2 years for the web phishing detection problem. Our algorithm achieves a average F1socre of 97%.