A neural network model for financial trend predicting

作者：Zhang, Xiumei^*; Ma, Chi; Yu, Xinmiao

来源：Cluster Computing, 2019, 22(2): S3487-S3494.

DOI：10.1007/s10586-018-2196-x

• Genetic Algorithm
• Fund trend predication
• Optimal Entropy
• Variation Function
• Fund net value

In this paper, we will put forward a hybrid model to predict the tendency of the fund, which is based on neural network combined with Genetic Algorithm (GA), Optimal Entropy (OE) and Variation Function. In China, financial market has great potential, but due to the intensity instability of the fund market. At present, the forecast about the trend of funds is still a not very accuracy. This model includes data pretreatment, input and output. Data preprocessing part is the time and characteristic alignment. Input part is that according to different characteristics, we use OE, GA and change function as of the input of neural network. Output part is using back propagation algorithm to get the final forecast. Because the input data has many different characteristics, in order to make full use of these characteristics, using the OE and GA of the global optimal feature selection can make the input model is more superior, which will help to improve the prediction accuracy of the proposed model. The experimental results show that using this hybrid model predicts china fund net value, the accuracy can reach 97.8%. This result is more accuracy when compare with China's existing institutions and financial website is more accurate.

• 出版日期2019-3
• 单位辽宁科技大学