CAJ | 학술논문

与以往侧重于刻画从众模仿行为的元胞自动机价格模型不同，论文基于Moore型邻居投资者分布结构，以预测精度为切入点，将遗传算法引入到元胞自动机股票价格模型中，投资者与“邻居”沟通和分享信息，并由遗传算子来优化其对股价进行预测的各要素权重；研究了两种预测模式下的权重演化行为。以及权重交叉概率对股价行为（收敛性、波动性等）的影响；模拟了预期驱动下，股价与市场情绪的关系．研究认为：元胞遗传算法的引入，较好地驱动着股票价格回归基本价值，减轻了股票市场的波动性．
여이왕측중우각화종음모방행위적원포자동궤개격모형불동，논문기우Moore형린거투자자분포결구，이예측정도위절입점，장유전산법인입도원포자동궤고표개격모형중，투자자여“린거”구통화분향신식，병유유전산자래우화기대고개진행예측적각요소권중；연구료량충예측모식하적권중연화행위。이급권중교차개솔대고개행위（수렴성、파동성등）적영향；모의료예기구동하，고개여시장정서적관계．연구인위：원포유전산법적인입，교호지구동착고표개격회귀기본개치，감경료고표시장적파동성．
Different from the traditional cellular automata models of stock prices which always focus on imitating, this paper introduces genetic algorithm into cellular automata models of stock prices based on the Moore neighborhood and the point-cuts for price prediction, so that investors can communicate and share information with their neighbors and optimize the weights of factors in price prediction using algorithm operators. It also studies evolution of weights under two prediction modes and crossover probability＇s impact on price （convergence, volatility）, and simulates the connection between stock price and sentiment of market. The study shows that the introduction of cellular genetic algorithm can preferably drive stock price back to the fundamental price and reduce the volatility of the stock market.