Process and system for developing a predictive model

作者: Matthias Kehder , David S. Dillon

DOI:

关键词:

摘要: The present invention relates to a computer implemented process for developing model which predicts the value of single dependent variable based on at least one independent (Figure 6, 100, 102, 104, 106, 108, 110, 112). comprises steps creating dataset containing plurality observations each and values variable, from original chromosomes comprising possible predictive model, quantitative fitness measure chromosome, new generation by selecting number upon measures, crossing selected cloning pure (standard) crossover technique, mutating crossed chromosomes. A system carrying out is also described.

参考文章(11)
Z. Michalewicz, C. Z. Janikow, GENOCOP: a genetic algorithm for numerical optimization problems with linear constraints Communications of The ACM. ,vol. 39, pp. 175- 201 ,(1996) , 10.1145/272682.272711
Cory J. Hoelting, Dale A. Schoenefeld, Roger L. Wainwright, Approximation techniques for variations of the p-median problem acm symposium on applied computing. pp. 293- 299 ,(1995) , 10.1145/315891.315997
E.F. Khor, K.C. Tan, M.L. Wang, T.H. Lee, Evolutionary algorithm with dynamic population size for multi-objective optimization international conference on industrial electronics control and instrumentation. ,vol. 4, pp. 2768- 2773 ,(2000) , 10.1109/IECON.2000.972436
G.G. Yen, Haiming Lu, Hierarchical genetic algorithm based neural network design 2000 IEEE Symposium on Combinations of Evolutionary Computation and Neural Networks. Proceedings of the First IEEE Symposium on Combinations of Evolutionary Computation and Neural Networks (Cat. No.00. pp. 168- 175 ,(2000) , 10.1109/ECNN.2000.886232
Aloke Guha, Steven A. Harp, Tariq Samad, Genetic synthesis of neural networks. ,(1990)