针对碳排放指标的复杂性与多样性,利用系统聚类对世界碳排放指标进行筛选;然后运用BP神经网络对世界碳排放量进行预测.结果表明运用系统聚类方法分析碳排放指标,简化了BP神经网络输入层,使网络训练得到较快实现,相比传统BP神经网络预测方法具有更高的精度,为碳排放预测及其它相关预测提供了一种新的、可供借鉴的方法. According to the complexity and diversity of carbon emission targets, the world carbon emission targets were screened by systematic clustering, and then the world carbon emissions were predicted by using BP neural network.The results showed that using the method of systematic clustering to analyze the carbon emission targets, The input layer of BP neural network can make the network training faster, which has higher accuracy than the traditional BP neural network prediction method, and provides a new and reference method for carbon emission prediction and other related prediction.