论文部分内容阅读
设计了一种基于粒子群优化的数据分类算法。新算法首先对数据样本预处理,利用粒子群优化算法通过训练数据进行分类规则的提取,根据提取得到的规则对数据进行分类识别。基于Bayes定理和随机状态转移过程对新算法的收敛性进行分析。通过对UCI数据集分类实验及遥感图像目标识别实验,验证了新算法是一种有效的分类方法。
A data classification algorithm based on particle swarm optimization was designed. The new algorithm firstly preprocesses the data sample, uses the particle swarm optimization algorithm to extract the classification rules through the training data, and classifies the data according to the extracted rules. The convergence of the new algorithm is analyzed based on the Bayes theorem and the stochastic state transition process. The experiment of UCI dataset classification experiment and remote sensing image recognition experiments verify that the new algorithm is an effective classification method.