Theoretical Analysis and Performance Improvement of Piecewise Linear Approximation and Statistical Learning Method of Competitive Associative Nets in Engineering Applications

2014 Fiscal Year Final Research Report

• Project/Area Number: 24500276
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Sensitivity informatics/Soft computing
• Research Institution: Kyushu Institute of Technology
• Principal Investigator: KUROGI Shuichi 九州工業大学, 工学(系)研究科(研究院), 教授 (40178124)
• Project Period (FY): 2012-04-01 – 2015-03-31
• Keywords: 競合連想ネット / 区分的線形近似 / 統計的学習 / 工学応用 / 理論解析 / 性能向上

Outline of Final Research Achievements

This research study has aimed to analyze and improve the performance of piecewise linear approximation and statistical learning method of competitive associative nets called CAN2 in engineering applications. Here the CAN2 is an artificial neural net for learning efficient piecewise linear approximation of nonlinear function. We have (1) analyzed the piecewise linear approximation and statistical learning method for performance improvement, (2) analyzed and improved multi-objective robust control of nonlinear plants, (3) proposed and analyzed a new method of text-prompted speaker identification, and (4) analyzed and improved range image processing methods. We have shown the effectiveness of the analysis and the methods improved in this research.

Free Research Field

知能制御工学