基于BIC测度和混合遗传算法的BNC结构学习

贝叶斯网络分类器(BNC)结构学习是一个NP难题。贪婪搜索(GS)算法是一种有效且准确性较高的结构学习算法,但贪婪搜索算法很容易陷入局部最优。标准遗传算法是一种全局搜索优化算法,它通过模拟生物种群的进化过程,得到全局最优解。但就其个体而言,个体局部解的质量无法保证,不具备局部寻优的能力。提出了将两种算法相结合,以贝叶斯信息标准(BIC)测度为评价函数,得到一种混合遗传算法,实现了它们的优势互补。实验表明:该算法优于单独利用GS算法进行Bayesian网络结构学习,从而说明该算法的正确性和有效性。

Structure Learning of BNC Based on BIC and Hybrid Genetic Algorithms

Structure learning of Bayesian networks classification is an NP hard problem.Greed search algorithm is an effective and high veracity method,but it is easy to get into the local best.Standard genetic algorithm is a global search optimal algorithm,which simulates the proceeding of natural evolution and can gets the global best.But its individual can't provide guarantee of getting the the local best.An algorithm is proposed to combine these two algorithms with BIC as evaluation function,which can get better effect.Experimental result shows that this algorithm is better than using GS algorithm only,it is accurate and effective.