| Popis: |
Accurate and reliable data is very important to the power system. In this paper, based on the analysis of GSA (Gap Statistic Algorithm) data mining technology, an improved clustering method is proposed for the shortcomings of the traditional method. The GSA method is an algorithm for determining the most appropriate number of clusters. The raw data is processed by BP neural network, and the preprocessed results are clustered by an improved clustering algorithm. Then use the elbow criterion to judge the optimal number of clusters. Compared with the original algorithm, it can overcome the defects in calculation speed and recognition accuracy, determine the number of clusters faster, greatly shorten the running time, and successfully identify the bad data in the power system. |
