Re-crime risk early warning mixed attribute data processing method, medium and equipment

Abstract

The invention discloses a re-crime risk early warning mixed attribute data processing method, medium and equipment, and the method comprises the steps: firstly obtaining data samples to form a data set, carrying out the preliminary data preprocessing of each sample in the data set, removing redundant items and missing items in the data set, and converting the data set into a coordinated data set; performing reduction processing on each condition attribute in the coordinated dataset, deleting redundant attributes in the coordinated dataset to obtain a dataset after attribute reduction, and finally clustering the dataset after attribute reduction. According to the method, missing attributes and redundant attributes existing in the data set can be effectively removed through data preprocessing and reduction processing, so that the dimension of the data can be effectively reduced, effective processing and analysis of the early warning data are realized, and through the data obtained based on the method, the classification accuracy of re-crime risk early warning can be higher, and the classification speed is higher.

Description

technical field [0001] The invention relates to the field of data preprocessing, in particular to a method, medium and equipment for processing mixed attribute data of re-crime risk early warning. Background technique [0002] The factors that lead to recidivism among special groups who have undergone prison reform are different, such as poor growth environment, deformed outlook on life, difficulty adapting to today's society after being released from prison, etc. And different recidivism factors have different motives and the degree of harm they bring to society. Therefore, according to the characteristics of different prisoners, using unsupervised clustering to divide the target population, and then discussing the factors that lead to recidivism and the degree of harm of each type of population, can make the prediction results more accurate and reduce the prediction algorithm. Prejudice against a particular group of people. [0003] However, due to the large scale of rec...

AI Technical Summary

Problems solved by technology

[0003] However, due to the large scale of recriminal risk early warning data and high sample dimensions, the existing models have high difficulty and low analysis ability in processing data with higher latitudes.

Moreover, the data contains both continuous attributes and categorical attributes. Currently, the cluster analysis of data in the existing technology is mainly for continuous attributes, but there are few techniques for cluster analysis of categorical attributes, and few clusters Analysis techniques can deal with continuous attributes and categorical attributes at the same time, and they are sensitive to the selection of initial cluster centers, which can easily cause the clustering results to fall into local optimum and cause the clustering effect to fluctuate greatly, and it is difficult to describe the samples and The dissimilarity between cluster centers and the differences between samples, without considering the different influences of the importance of different conditional attributes on the results, etc.

Images