Practical algorithms for incremental growth.

MacLeod, Christopher

Abstract

This report considers some of the practical issues involved in the implementation of Incremental Growth or Incremental Evolution algorithms as outlined in the paper: Incremental Growth in Modular Neural Networks (doi:10.1016/j.engappai.2008.11.002), originally published in the journal Engineering Applications of Artificial Intelligence and the article: Minds for Robots, published in the magazine Electronics World. These algorithms allow a Neural Network or similar system to grow, piece by piece, in a controlled manner. The sections below consider the data structures, algorithms and programming techniques which can be used and also addresses unit functionality and possibilities for interesting further work.

Citation

MACLEOD, C. 2010. Practical algorithms for incremental growth. Aberdeen: Robert Gordon University.

• Report Type: Research Report
• Publication Date: Dec 31, 2010
• Deposit Date: Mar 8, 2010
• Publicly Available Date: Mar 8, 2010
• Publisher: Robert Gordon University
• Keywords: Algorithms; Incremental Growth
• Public URL: http://hdl.handle.net/10059/471
• Contract Date: Mar 8, 2010