The combination stretching function technique with simulated annealing algorithm for global optimization

The Application of Pattern Recognition in Electrofacies Analysis

“…In (3), selecting the one with the minimum sum of squares of deviations as the improved initial clustering center can reduce the randomness brought by random selection. In (2), to eliminate the influence of error point, the modified Kmeans algorithm (KmeansMod) is adopted. KmeansMod has the following modification based on the standard K-means: when the standard K-means algorithm is completed, the data point contained in each clustering will be checked.…”

“…There are two ways to eliminate gibberish, simplify control methods, and reduce calculated amount: (1) principal component analysis. (2) Select logging items manually. The extracted logging items will be recorded in Table stdlogdata as the data source for clustering analysis.…”

“…The main drawback of the KM algorithm is that the cluster result is sensitive to the selection of the initial cluster centers and may converge to the local optima. At present, the development tendency of clustering method is to find a global optimal solution in combination with the global optimization methods like simulated annealing, particle swarm, and other local methods like K-means [2][3][4][5][6]. Pelleg and Moore [7] proposed an algorithm which can automatically determine the optimal number of clusters during clustering.…”

Performance Analysis of Bi-Attempted Based Optimization Algorithm on Numerical Optimization Functions

“…Eldeki soruna bağlı olarak, tüm algoritmalar bazı iyileştirmeler gerektirmektedir. Bu nedenle, yerel arama yöntemi sezgisel algoritmaların performansını arttırmak için iyi bilinen ve sık kullanılan yöntemlerden biridir [36][37][38][39][40][41]. Literatürde çok çeşitli meta-sezgisel algoritmalar bulunmaktadır.…”