Quantum clustering algorithm improvement method oriented to classification attribute data

Abstract

According to the classification attribute data-oriented quantum clustering algorithm improvement method provided by the invention, the potential energy of the data points is calculated by using the quantum potential energy function according to the characteristics of the classification attribute data, and the potential classification problem of the classification data is solved. The lowest point of potential energy is used as the center of clustering, the scale parameter of the wave function is calculated by using the kernel width adjustment parameter, and the clustering range of the classification attribute data is solved. Furthermore, the influence of data attributes on a classification result is considered, a traditional Euclidean distance is utilized, an estimation formula of a kernel width adjustment parameter is given, and the calculation problem of an internal relation between the number of samples and dimensions is solved. The method has good calculation accuracy for classification attribute data.

Description

technical field [0001] The invention belongs to the field of data mining. Background technique [0002] With the continuous development of navigation information technology, the resulting information data contains rich ship information, but how to make good use of these massive information data to escort ships has become the biggest difficulty. In order to get rid of the dilemma of "rich data, poor knowledge", we began to study how to mine useful and usable information from massive data. The ship's navigation trajectory is the most intuitive embodiment of the ship's behavior. By analyzing the ship's trajectory, on the one hand, the sea state information can be obtained, and on the other hand, the ship's navigation behavior can be detected. Through the extraction, clustering, modeling and anomaly detection of ship trajectories, the description and analysis of ship behavior and real-time anomaly detection can be realized, and warnings can be made at the early stage of ship ab...

AI Technical Summary

Problems solved by technology

One is to find the lowest point of effective potential energy by adjusting the parameters of the potential energy calculation formula, but ignore the influence of distance on the clustering effect. Due to the huge amount of data and the complex potential energy calculation formula, this type of improved quantum clustering Such algorithms have high complexity and long running time, making it difficult to meet the needs of practical applications

Another type of improved quantum clustering algorithm mainly focuses on data preprocessing before clustering, and reduces data running time through early data preprocessing, but during data preprocessing, effective information may be lost

Images