A genetic spectral clustering algorithm

作者：Wang Huiqing^*; Chen Junjie; Guo Kai

来源：Journal of Computational Information Systems, 2011, 7(9): 3245-3252.

• Artificial data
• Data sets
• Dimension reduction
• Genetic K-means algorithm
• Global search ability
• Input dimensions
• Iterative Optimization
• k-Means algorithm
• K-means clustering
• Local optimal solution
• Machine-learning
• Optimal solutions
• Sample space
• Spectral clustering
• Spectral graph theory
• Stable clusters

As a novel clustering algorithm, spectral clustering is applied in machine learning extensively. Spectral clustering is built upon spectral graph theory, and has the ability to process the clustering of non-convex sample spaces. Most of the existing spectral clustering algorithms are based on k-means algorithm, and k-means algorithm uses the iterative optimization method to find the optimal solution, which is easy to prematurely converge to the local optimal solution. Combined with the global search ability of genetic algorithm, a genetic spectral clustering algorithm is proposed. Compared with the original spectral clustering and k-means clustering analysis based on genetic algorithm, the suggested algorithm reduces the input dimension of the clustering algorithm using dimension reduction of spectral clustering, and replaces the traditional k-means algorithm by genetic k-means algorithm. The experiments show that the suggested algorithm obtains the stable cluster centers, and effectively improves the clustering performance on both artificial data and UCI datasets, which validate the stability and effectiveness of the suggested algorithm.

• 出版日期2011
• 单位太原理工大学