CAJ | 학술논문

在数据采集过程中结合网格聚类算法提高计算效率，为了保存采样数据的分布特点引入权值。根据类别中心密度高、权值大的特征采用寻找连通分量的方法初步确定聚类中心，在此基础上结合自适应免疫算法，动态地确定聚类中心及其类别数。进而使 FCM 算法跳出局部最优，最大可能地得到全局最优解。
재수거채집과정중결합망격취류산법제고계산효솔，위료보존채양수거적분포특점인입권치。근거유별중심밀도고、권치대적특정채용심조련통분량적방법초보학정취류중심，재차기출상결합자괄응면역산법，동태지학정취류중심급기유별수。진이사 FCM 산법도출국부최우，최대가능지득도전국최우해。
The grid clustering algorithm is combined to improve the computational efficiency in the data collection process , and the right values are introduced to preserve the distribution characteristics of the sampled data. Depending on the characters of large weight and density in the category center, methods of finding connected components are adopted to initially identify cluster centers. On this basis, the adaptive immune algorithm is combined to dynamically determine the cluster centers and the number of categories. Thereby the FCM algorithm is able to escape from local optima, and the maximum possible to get a global optimal solution.