Title :
Clonal operator and antibody clone algorithms
Author :
Du, Hai-feng ; Jiao, Li-Cheng ; Wang, Sun-an
Author_Institution :
Key Lab. of Radar Signal Process., Xidian Univ., Xi´´an, China
Abstract :
Based on clonal selection theory, the main mechanisms of a clone in the immune system, which are explored in the field of artificial intelligence, are analyzed. An artificial immune system algorithm, antibody clone algorithm, is put forward. Compared with an improved gene algorithm, the new algorithm is shown to be an evolutionary strategy capable of solving complex machine learning tasks, like multi-modal optimization. Using Markov chain theories, it is proved that the antibody clone algorithm is convergent.
Keywords :
Markov processes; convergence; evolutionary computation; learning (artificial intelligence); optimisation; Markov chain; antibody clone algorithms; artificial immune system algorithm; artificial intelligence; clonal operator; clonal selection theory; complex machine learning tasks; convergent algorithm; evolutionary strategy; multi-modal optimization; Artificial immune systems; Artificial intelligence; Biology computing; Cloning; Evolution (biology); Genetic mutations; Immune system; Machine learning algorithms; Signal processing algorithms; Stochastic processes;
Conference_Titel :
Machine Learning and Cybernetics, 2002. Proceedings. 2002 International Conference on
Print_ISBN :
0-7803-7508-4
DOI :
10.1109/ICMLC.2002.1176807
