关键词: 群体状态控制, 优化算法, 函数极值, 模式分类

Abstract:

In the era of big data, the amount of data for optimization problems is constantly increasing, and the dimension is constantly increasing, and it has more complex multi-mode properties. Therefore, the traditional optimization algorithms can not meet the requirements of people. Inspired by the phenomenon that the state of matter changes tends to be stable, and according to universal gravitation and four particle motion modes, this paper proposes a novel Colon State Adaptive Optimization algorithm based on Gravitation（G-CSAO）. In the function optimization experiment, the proposed algorithm has higher precision of optimization, and is not easy to fall into local optimal value, and can better cope with multi-peak problems and high-dimensional problems. The proposed model is combined with the minimum distance criterion in the pattern classification problem. The experimental results of UCI dataset show that the proposed model has better classification performance and application prospects.

Key words: colon state control, optimization algorithm, function optimization problem, pattern classification