Research on Network Active Detection Technology of Power Monitoring System Based on Machine Learning

Abstract

There are still many security risks in the current power monitoring system network, and lack of independent learning ability. Based on this, a network learning based power detection system network active detection technology is proposed. The scheme analyzes and processes the data, uses feature engineering to process the features, and improves the common filtering method by using the feature correlation matrix in feature selection. Model training and prediction based on threshold-enhanced k-means semi-supervised algorithm. The experimental results show that the program has a certain accuracy rate, has certain self-learning ability, and has certain feasibility.