應用智慧型基因演算法(IGA)於統測成績及大學生身高的建模與預測

Intelligent genetic algorithm (IGA) for modeling and prediction of Math achievement and student’s height

陳彥佑(Yan-You Chen)

指導教授 ： 何信璋

國立虎尾科技大學/工程學院/自動化工程研究所/碩士(2010年)

摘要

近年來基因演算法已經廣泛的應用於多方面，包含自動控制、系統優化、生醫研究等。而本文使用智慧型基因演算法(IGA)配合模糊神經網路(FNN)的方式，來預測數學成績及大學生身高的數據。IGA是GA配合直交實驗設計的智慧型演算法，可以有效的推理出近似最佳解。

本文的兩組數據的建模是使用問卷調查的方式，依據調查結果來做統計，主要有100組訓練數據及10組的測試數據，使用FNN架構建模，模型參數都是126個。而由實驗結果得知，以FNN的架構配合IGA來搜尋最好的FNN模型參數，能夠預測出誤差較小的結果，並且會優於MATLAB的ANFIS方法所預測出來的結果。

關鍵字

智慧型基因演算法(IGA) ； 模糊神經網路(FNN) ； 直交實驗設計 ； 預測

並列摘要

Genetic algorithms have been widely used in many ways, including automatic control, system optimization, biomedical research in recent years. The article uses an intelligent genetic algorithm (IGA) with the fuzzy neural network (FNN) approach to predict mathematics achievement and the students height data. IGA is a GA with orthogonal experimental design of intelligent algorithms, it can effectively be inferred that near optimal solution.

This article uses two sets of data modeling is the way of questionnaire. Do statistics based on survey results, the main group of 100 training data and 10 test data, using the FNN architecture modeling, model parameters are 126. By experimental results, the FNN architecture with IGA to search for the best FNN model parameters, can predict the result of small errors, and outperforms the ANFIS methods by MATLAB to forecast results.

並列關鍵字

Intelligent Genetic Algorithm (IGA) ； Fuzzy Neural Networks (FNN) ； Orthogonal Experimental Design(OED) ； Prediction

參考文獻

[6] 黃忠雄, 應用智慧型基因演算法(IGA)於造血幹細胞(HSCs)與粗糙度的建模與預測, 虎尾科技大學機械與機電工程研究所碩士論文, 2008.

[7] 梁宏維, 智慧型基因演算法在粗糙度的建模與預測, 虎尾科技大學機械與機電工程研究所, 2008.

[1] Shinn-Ying Ho, Li-Sun Shu, Jian-Hung Chen, “Intelligent Evolutionary Algorithms for Large Parameter Optimization Problems”, Evolutionary Computation, IEEE, pp. 522 -541,2004

[2] S.-Y. Ho, S.-J. Ho, and T.-K. Chen, "Design of High Performance Fuzzy Controllers Using Flexible Parameterized Membership Functions and Intelligent Genetic Algorithms," International Journal JSME Series C, vol. 46, no. 1, pp. 252-262, 2003.

[3] S.-Y. Ho, C,-C. Liu, and S. Liu, "Design of an Optimal Nearest Neighbor Classifier Using an Intelligent Genetic Algorithm," Pattern Recognition Letters, vol. 23, no. 13, pp. 1495-1503, May 2002.