计算机工程与设计
計算機工程與設計
계산궤공정여설계
COMPUTER ENGINEERING AND DESIGN
2009年
21期
4945-4948
,共4页
股票价格预测%隐马尔可夫模型%隐马尔可夫模型优化%粒子群优化算法%人工神经网络
股票價格預測%隱馬爾可伕模型%隱馬爾可伕模型優化%粒子群優化算法%人工神經網絡
고표개격예측%은마이가부모형%은마이가부모형우화%입자군우화산법%인공신경망락
stock market forecasting%hidden Markov model%hidden Markov model optimize%artificial neural network%particle swarm optimization
提出了一种用于股票价格预测的人工神经网络(ANN),隐马尔可夫模型(HMM)和粒子群优化算法(PSO)的组合模型-APHMM模型.在APHMM模型中,ANN算法将股票的每日开盘价、最高价、最低价与收盘价转换为相互独立的量并作为HMM的输入.然后,利用PSO算法对HMM的参数初始值进行优化,并用Baum-Welch算法进行参数训练.经过训练后的HMM在历史数据中找出一组与今天股票的上述4个指标模式最相似数据,加权平均计算每个数据与它后一天的收盘价格差,则今天的股票收盘价加上这个加权平均价格差便为预测的股票收盘价.实验结果表明,APHMM模型具有良好的预测性能.
提齣瞭一種用于股票價格預測的人工神經網絡(ANN),隱馬爾可伕模型(HMM)和粒子群優化算法(PSO)的組閤模型-APHMM模型.在APHMM模型中,ANN算法將股票的每日開盤價、最高價、最低價與收盤價轉換為相互獨立的量併作為HMM的輸入.然後,利用PSO算法對HMM的參數初始值進行優化,併用Baum-Welch算法進行參數訓練.經過訓練後的HMM在歷史數據中找齣一組與今天股票的上述4箇指標模式最相似數據,加權平均計算每箇數據與它後一天的收盤價格差,則今天的股票收盤價加上這箇加權平均價格差便為預測的股票收盤價.實驗結果錶明,APHMM模型具有良好的預測性能.
제출료일충용우고표개격예측적인공신경망락(ANN),은마이가부모형(HMM)화입자군우화산법(PSO)적조합모형-APHMM모형.재APHMM모형중,ANN산법장고표적매일개반개、최고개、최저개여수반개전환위상호독립적량병작위HMM적수입.연후,이용PSO산법대HMM적삼수초시치진행우화,병용Baum-Welch산법진행삼수훈련.경과훈련후적HMM재역사수거중조출일조여금천고표적상술4개지표모식최상사수거,가권평균계산매개수거여타후일천적수반개격차,칙금천적고표수반개가상저개가권평균개격차편위예측적고표수반개.실험결과표명,APHMM모형구유량호적예측성능.
A fusion model APHMM is proposed by combining the hidden Markov model (HMM), artificial neural networks (ANN) and particle swarm optimization (PSO) to forecast financial market behavior. In APHMM, use ANN to transform the daily stock price into independent sets of values and become input to HMM. Then draw on PSO to optimize the initial parameters of HMM. The trained HMM is used to identify and locate similar patterns in the historical data. The price differences between the matched days and the respective next day are calculated. Finally, a weighted average of the price differences of similar patterns is obtained to prepare a forecast for the required next day. Forecasts are obtained for a number of securities that show APHMM is feasible.
