基于特征选择的GDP预测研究

Research on GDP Forecasting Based on Feature Selection

做好国内生产总值(GDP)预测对地方国民经济发展规划及制定各行各业的发展计划具有指导性作用。由此提出了一种基于特征选择算法的GDP预测模型,该模型主要由特征选择模块和预测模块组成。特征选择模块被设计为只选择最相关的特征,通过计算系统属性与特征属性之间的关联,只有相关性最高的特征属性才能被保留。然后,将选取的特征属性作为输入数据进入预测模块进行预测。以安徽省的生产总值(GDP)为案例,对预测模型进行了验证。结果表明,通过利用提出的特征选择方法选择有用的特征,对GDP进行预测可显著提高预测模型的准确性和有效性。

The forecast of gross domestic product(GDP)plays a guiding role in the development plan of local national economy and the development plan of every industry.Therefore,a GDP prediction model based on feature selection algorithm is proposed,which is mainly composed of feature selection module and prediction module.The feature selection module is designed to select only the most relevant features.By calculating the correlation between system attributes and feature attributes,only the feature attributes with the highest correlation can be retained.Then,the selected feature attributes are entered into the prediction module as input data for prediction.Taking the GDP of Anhui Province as an example,the forecasting model is verified.The results show that by using the proposed feature selection method to select useful features,the accuracy and effectiveness of GDP prediction model can be significantly improved.