Sea area anomaly monitoring method and device based on improved K-means clustering algorithm model

The embodiment of the invention relates to a sea area anomaly monitoring method and device based on an improved K-means clustering algorithm model, and the method comprises the steps: obtaining data which is obtained by a data collection device or is extracted from a data set inputted by a user; carrying out minimum correlation degree judgment on the extracted data, dividing samples into a plurality of categories through an internal relationship among the data, and constructing a data set Dataset-1 based on an original sample range; and a pre-training prompt model is introduced, parameter standardization and abnormal point filtering are continuously performed on the data, global optimization under a local optimization condition is obtained, and accurate judgment of sea area abnormal sample points is realized. According to the sea area abnormal value forecasting method based on the improved mathematical model, the purposes of correcting the sea area condition error reporting rate and improving the abnormal mon