从时序数据中查找异常数据,需要将每一个时间快照的网络数据提取网络特征,

To find abnormal data from time series data, it is necessary to extract network features from the network data of each time snapshot, and use the similarity evaluation method to detect abnormal data for adjacent data, such as the number of nodes and the number of connecting edges that can be used for each time snapshot. The average clustering coefficient of each time snapshot can also be extracted to analyze the time points with large changes.

To find abnormal data from time series data, it is necessary to extract the network characteristics from the network data of each time snapshot, and the adjacent data use the similarity evaluation method to detect abnormal data. For example, the change of the number of nodes and connecting edges of each time snapshot can be used, and the average agglomeration coefficient of each time snapshot can also be extracted to analyze the time points with large changes

To find abnormal data from time series data, it is necessary to extract the network characteristics from the network data of each time snapshot, and to detect abnormal data from the adjacent data by using the evaluation method of similarity. For example, the change of the number of nodes and the number of connecting edges in each time snapshot can be used, and the average aggregation coefficient of each time snapshot can also be extracted, and the time points with large changes can be analyzed.