摘要: In this paper, we propose a hybrid neurogenetic system for stock trading. A recurrent neural network (NN) having one hidden layer is used the prediction model. The input features are generated from number of technical indicators being by financial experts. genetic algorithm (GA) optimizes NN's weights under 2-D encoding and crossover. We devised context-based ensemble method NNs which dynamically changes on basis test day's context. To reduce time in processing mass data, parallelized GA Linux cluster using message passing interface. tested proposed with 36 companies NYSE NASDAQ 13 years 1992 to 2004. showed notable improvement average over buy-and-hold strategy further improved results. also observed that some were more predictable than others, implies can be portfolio construction
acm.org
sci-hub.se 参考文章(39)
Perry J. Kaufman, Trading Systems and Methods ,(1997)
Nils John Nilsson, Learning machines: foundations of trainable pattern-classifying systems ,(1965)
Andrew B. Kahng, Byung Ro Moon, Toward More Powerful Recombinations international conference on genetic algorithms. pp. 96- 103 ,(1995)
Aloke Guha, Steven Alex Harp, Tariq Samad, Towards the genetic synthesis of neural network international conference on genetic algorithms. pp. 360- 369 ,(1989)
David E. Goldberg, Genetic algorithms in search, optimization, and machine learning ,(1988)
Byung Ro Moon, Jung Hwan Kim, Neuron Reordering For Better Neuro-genetic Hybrids genetic and evolutionary computation conference. pp. 407- 414 ,(2002)
H. Iba, N. Nikolaev, Genetic programming polynomial models of financial data series congress on evolutionary computation. ,vol. 2, pp. 1459- 1466 ,(2000) , 10.1109/CEC.2000.870826
Eugene F. Fama, Marshall E. Blume, Filter Rules and Stock-Market Trading The Journal of Business. ,vol. 39, pp. 226- 241 ,(1966) , 10.1086/294849
Hendrik Bessembinder, Kalok Chan, The profitability of technical trading rules in the Asian stock markets Pacific-basin Finance Journal. ,vol. 3, pp. 257- 284 ,(1995) , 10.1016/0927-538X(95)00002-3
Iebeling Kaastra, Milton Boyd, Designing a neural network for forecasting financial and economic time series Neurocomputing. ,vol. 10, pp. 215- 236 ,(1996) , 10.1016/0925-2312(95)00039-9