论文部分内容阅读
本文利用支持向量机(SVM)机器学习算法,融合宏观经济数据、上海现货铜价格、LME期铜价格和美元汇率来研究预测上海期货交易所铜期货价格。通过上海期货交易所cu1402合约210天交易数据和同期相关经济数据,来预测铜期货每日最高价、最低价、开盘价和收盘价。模型结果显示,对上海期货交易所铜期货cu1402合约价格预测存在一个25天左右的累计平均误差最小的规律。另外,也对影响沪铜期货每日最高价、最低价、开盘价和收盘价的经济数据分别作了分析,其中沪铜期货价格的历史数据、铜现货价格和LME铜期货价格都对沪铜四种价格预测存在重要影响作用。
In this paper, SVM machine learning algorithm, the integration of macroeconomic data, the Shanghai spot copper price, the LME copper price and the US dollar exchange rate to study the prediction of the Shanghai Futures Exchange copper futures prices. Through the Shanghai Futures Exchange cu1402 contract 210 days of transaction data and related economic data over the same period, to predict the daily maximum copper futures, the lowest price, opening price and closing price. The results of the model show that there is a law of the smallest cumulative average error of about 25 days in predicting the contract price of cu1402 on the Shanghai Futures Exchange. In addition, the daily prices, the lowest prices, the opening prices and the closing prices of Shanghai Copper futures were also analyzed separately. The historical data of the Shanghai copper futures prices, the spot copper prices and the LME copper futures prices also analyzed the price of Shanghai Copper Four kinds of price forecasts have a significant impact.