高质量发展区域经济数据异常趋势智能预测研究

Research on Intelligent Prediction of Abnormal Trend of Data in High Quality Development Regional Economic

为了把握经济发展规律,合理制定经济政策,开展了高质量发展区域经济中的数据异常趋势智能预测研究。遵循全面性、相关性、周期性等准则,从高质量供给、需求、发展效率、对外开放等方面采集地区经济数据;通过数据集成、标准化与归一化等操作完成数据预处理,确保数据具有一致性特征,并选择具有代表性的特征向量,采用主成分分析法实现异常数据特征提取;引入信息熵概念,确定样本集合,通过递归方式对所有数据分类,构建决策树;建立时间序列矩阵,计算候选数据对的信息熵与信息增益量,结合时间规整距离实现正常与异常趋势的预测。仿真实验证明,该方法预测的趋势与实际发展趋势吻合度较高,且减少了预测时间。

In order to grasp the law of economic development and formulate economic policies reasonably,the intelligent prediction of abnormal trends of high-quality regional economic data is put forward.Following the principles of comprehensiveness,relevance and periodicity,we collect regional economic data from the aspects of high-quality supply,demand,development efficiency and opening policies.Through data integration,standardization and normalization,data preprocessing is completed to ensure that the data have consistent features,and representative feature vectors are selected to extract abnormal data features by principal component analysis.We introduce the concept of information entropy,determine the sample set,classify all the data by recursion,and build a decision tree.The time series matrix is established,the information entropy and information gain of candidate data pairs are calculated,and the prediction of normal and abnormal trends is realized by combining the regular time distance.Simulation experiments show that the predicted trend of this method is in good agreement with the actual development trend,and the prediction time is reduced.