Feature Extraction of Marine Water Pollution Based on Data Mining

Abstract

The ocean occupies more than two-thirds of the earth’s area, providing a lot of oxygen and materials for human survival and development. However, with human activities, a large number of sewage, plastic bags, and other wastes are discharged into the ocean, and the problem of marine water pollution has become a hot topic in the world. In order to extract the characteristics of marine water pollution, this study proposed K-means clustering technology based on cosine distance and discrimination to study the polluted water. In this study, the polygonal area method combined with six parameters of water quality is used to analyze the marine water body anomalies, so as to realize the rapid and real-time monitoring of marine water body anomalies. At the same time, the cosine distance method and discrimination are used to classify marine water pollutants, so as to improve the classification accuracy. The results show that the detection rate of water quality anomalies is more than 88.2%, and the overall classification accuracy of water pollution is 96.3%, which proves the effectiveness of the method. It is hoped that this study can provide timely and effective data support for the detection of marine water bodies.