Research on similar time series data search algorithm based on dimension-by-dimension strategy

Author:

Wang Dacheng1,Teng Qizhi1,Jiao Yue1,Zhang Ling1

Affiliation:

1. Chinese Academy of Sciences

Abstract

Abstract The CS search algorithm is a new type of bionic optimization algorithm, which is simple and efficient, and has been successfully used in classical theoretical research and engineering applications. The algorithm uses the overall update evaluation solution to solve the continuous function optimization problem. Due to the phenomenon of mutual dimensional interference, the local refinement ability and convergence speed of the algorithm will be affected to some extent. This paper introduces the evaluation strategy of one-dimensional update and rewrites the step formula of the random walk component so that the algorithm makes full use of one-dimensional evolution information. It strengthens the local search, improves the convergence speed and the quality of the solution, and gives experimental results. The calculation results of different dimensions show that: as the dimension increases, the dimension-by-dimension strategy improves the convergence speed and solution quality. Compared with the related improved CS algorithm and other evolutionary algorithms, the improved algorithm has certain competitiveness in solving multi-dimensional function optimization problems.

Publisher

Research Square Platform LLC

Reference10 articles.

1. A Computationally Efficient Evolutionary Algorithm for Real-Parameter Optimization;Deb K;Evol Comput,2002

2. Completely Derandomized Self-Adaptation in Evolution Strategies;Hansen N;Evol Comput,2001

3. Hui Y, Guozhen Z (2012) Symbol Method for Fast Similarity Search in Large Time Series Flight Databases. In: Proceedings of the 2012 International Conference on Convergence Computer Technology. IEEE Computer Society, USA, pp 7–10

4. Parameter estimation for chaotic systems using the cuckoo search algorithm with an orthogonal learning method;Li X;Chin Phys B,2012

5. Enhancing the performance of cuckoo search algorithm using orthogonal learning method;Li X;Neural Comput Appl,2014

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3