An Efficient Optimization Algorithm for Super High Dimensional Numerical Function Inspired by Cellular Differentiation

• Main Authors: Yanjiang Wang, Chengna Yuan, Hui Li, Yujuan Qi
• Format: Article
• Language: English
• Published: IFSA Publishing, S.L. 2013-11-01
• Series: Sensors & Transducers
• Subjects: Cellular differentiation optimization algorithm (CDOA); Bio-inspired; Complex function with super high dimensions; Numerical function optimization.
• Online Access: http://www.sensorsportal.com/HTML/DIGEST/november_2013/PDF_vol_159/P_1550.pdf

Inspired by the cellular differentiation behaviors, a new biomimetic optimization algorithm, cellular differentiation optimization algorithm (CDOA), is proposed. First, a certain number of cells are randomly distributed in the search space in which each cell represents a solution. Then, several cellular differentiation behaviors such as division, growth, migration, adhesion and apoptosis are exhibited for finding the optimal solution according to the activity value of a cell. The proposed algorithm is applied to several benchmark complex functions optimization with 20-1000 dimensions. Experimental results show that the proposed cellular differentiation optimization algorithm can converge to the optimum of complex numerical functions with super high dimensions rapidly in spite of its simple procedure and effortless implementation.