GAQPR and its application in discovering frequent structures in time series

Autor: Zhuang Zhen-quan, Yang Jun-an, Li Bin
Rok vydání: 2003
Předmět:
Zdroj: International Conference on Neural Networks and Signal Processing, 2003. Proceedings of the 2003.
DOI: 10.1109/icnnsp.2003.1279293
Popis: A new genetic algorithm based on the quantum probability representation (GAQPR) is proposed, in which each individual evolves independently; a new crossover operator is designed to integrate searching processes of multiple individuals into a more efficient global searching process; a new mutation operator is also proposed. The algorithm is used to discover frequent structures in time series, experiment results show that the GAQPR is efficient for the complex multi-peak searching problem.
Databáze: OpenAIRE