Data Anomaly Detection Based on Isolation Forest Algorithm
Liang Zhang, Lingyun Liu · 2022
Compared with other anomaly detection methods, the traditional isolation forest algorithm improves the execution efficiency, but it still is time-consuming. Thus, an improved isolation forest algorithm was proposed to solve the problem. Specifically, we analyzed the user access data stream of the website platform through data preprocessing. Then PCA was used to reduce the feature dimension, and a parallel processing approach based on an isolation forest algorithm was used to detect the data. The detection efficiency of the proposed method is verified through the standard simulation data set.