摘要
本文以沪深证券市场的实时分时数据为基础,应用神经网络技术对证券市场的八种经典分时形态进行了动态分割预处理和模式识别、预测,实验表明上述方法具有良好的稳定性和可靠性,并抗噪能力强且准确率较高。
In this paper,the true interday data of stock mar ket in Shanghai and Shenzhen are analyzed.Meant for eight kinds of classical patterns of stock mar ket,the paper,based on neural network technology,dis-cuss es one new al gorithm for dynamic pat tern an tici -pated segmentation,pat tern recognition and fore cast -ing.Ex periments in di cate these methods have good stability and reliability.The al gorithm sys tem is su perior in noisy im ages and accu rate detection and recogni -tion.
出处
《地质技术经济管理》
2003年第1期41-43,共3页
Geological Technoeconomic Management