Traffic analysis and prediction method based on big data

The invention provides a traffic analysis and prediction method based on big data. The method specifically comprises the following steps: collecting to-be-predicted traffic data; the data are input into a traffic characteristic analysis module for time sequence characteristic analysis to obtain traffic characteristic data; obtaining potential associated data, and inputting the potential associated data into an associated feature analysis module after data processing to obtain associated feature data; inputting the second prediction data set and the associated feature data into a traffic prediction module for traffic prediction to obtain a traffic prediction result; by means of the method, the prediction result oriented to the network game scene can be more accurate, and the real-time performance of traffic prediction is high. 本发明提出了基于大数据的流量分析预测方法，具体包括：收集待预测流量数据；输入流量特征分析模块进行时序特征分析，得到流量特征数据；获取潜在关联数据，经数据处理后输入关联特征分析模块，得到关联特征数据；将第二预测数据集及关联特征数据输入流量预测模块进行流量预测，得到流量预测结果；借由上述方法，可实现面向网络游戏场景的预测结果更为准确，流量预测的实时性强。