PBIL进化算法在自动组卷系统中的应用  

Application of PBIL Algorithm in Automatic Test Paper Construction

作　　者：楼玉萍[1] 金炳尧[1] 骆红波[1] 

机构地区：[1]浙江师范大学信息学院

出　　处：《计算机技术与发展》2006年第6期80-82,共3页Computer Technology and Development

摘　　要：进化计算是一种搜索方法,广泛用于求解各类优化问题。PBIL算法将进化获得的知识———学习概率用以指导后代的产生,使搜索更具方向性,因而往往能取得更好的效果。自动组卷问题是一个典型的组合优化问题。文中针对PBIL算法的特点,设计了一个自动组卷求解方案,并用实验数据进行计算。结果表明:该算法计算速度快、稳定性好,尤其是在约束条件比较多的情况下,显示出算法的高适应性,是解决组卷问题较为理想的算法。Evolutionary computation is a search method,which has been applied in solving optimization problems. PBIL algorithm instructs the generation of offspring through knowledge acquired from evolutionary, namely, learning probability, and makes the searching in the right direction. So can get the better result. Automatic test paper construction is a typical optimization problem. The paper brings forward an automatic test paper construetion solution through PBIL algorlthm, and evolutionary computation is realized by experiment data. The result shows that the algorithm has high - speed and good- stability. Especially under the condition with many limits, the algorithm shows that it is more adaptive and it is an ideal algorithm for solving the problem of paper organization.

关 键 词：进化计算 PBIL算法 组卷问题 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]