Re-crime risk prediction method based on reinforcement learning, medium and computing device

A technology of risk prediction and reinforcement learning, applied in the field of re-crime risk prediction based on reinforcement learning, can solve problems such as difficult re-crime prediction

Inactive Publication Date: 2020-10-13
SOUTH CHINA AGRI UNIV
View PDF0 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Therefore, it is difficult to accurately predict the recidivism of relevant personnel and give early warning in time only by using such methods.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Re-crime risk prediction method based on reinforcement learning, medium and computing device
  • Re-crime risk prediction method based on reinforcement learning, medium and computing device
  • Re-crime risk prediction method based on reinforcement learning, medium and computing device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0086] This embodiment discloses a method for predicting the risk of recidivism based on reinforcement learning. This method can target relevant personnel with criminal history and then predict criminal behavior, so as to be able to conduct targeted monitoring and intervention for these personnel to reduce The impact of re-crime on society. The steps of the re-crime risk prediction method in this embodiment are as follows figure 1 As shown in, including:

[0087] S1. Obtain training samples to form a training sample set; training samples include persons with criminal records and re-criminal behaviors and persons with criminal records and no re-criminal behaviors.

[0088] As shown in Table 1, suppose the following training samples are included in the training sample set, and the data of each training sample is as follows:

[0089] Table 1

[0090] Numbering gender age Type of crime Whether to commit a crime again a1 male19 violence Yes a2 male27 theft no a3 male45 robbery Yes a...

Embodiment 2

[0203] The storage medium includes a processor and a memory for storing an executable program for the processor, wherein the processor executes the program stored in the memory to implement the recidivism risk based on reinforcement learning described in Embodiment 1. The forecast method is as follows:

[0204] Obtain training samples to form a training sample set; training samples include persons with criminal records and re-criminal behaviors and persons with criminal records and no re-criminal behaviors;

[0205] According to the continuous attribute and sub-type attribute of each training sample in the training sample set, cluster each training sample, define the number of clusters obtained as N, and N is a constant, that is, all training samples in the training sample set are clustered as N class;

[0206] For the N classes obtained by clustering, construct corresponding N neural networks respectively;

[0207] Input the continuous attribute of each training sample and the sub-t...

Embodiment 3

[0212] This embodiment discloses a computing device that stores a program, and when the program is executed by a processor, the method for re-crime risk prediction based on reinforcement learning described in Embodiment 1 is implemented as follows:

[0213] Obtain training samples to form a training sample set; training samples include persons with criminal records and re-criminal acts and persons with criminal records and no re-criminal acts;

[0214] According to the continuous attribute and sub-type attribute of each training sample in the training sample set, cluster each training sample, define the number of clusters obtained as N, and N is a constant, that is, all training samples in the training sample set are clustered as N class;

[0215] For the N classes obtained by clustering, construct corresponding N neural networks;

[0216] Input the continuous attribute of each training sample and the sub-type attribute after one-hot encoding into the neural network corresponding to t...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a re-crime risk prediction method based on reinforcement learning, a medium and computing equipment, and the method comprises the steps: firstly constructing a training sampleset, and carrying out the clustering of the training sample set; for N classes obtained by clustering, respectively constructing N BP neural networks; inputting each training sample attribute into a corresponding BP neural network, and training the BP neural network to obtain a re-crime risk prediction model; for a test sample needing to predict a re-crime risk, calculating the distance between the test sample and each clustering center; selecting a clustering center with the smallest distance from the test sample; and taking the trained neural network corresponding to the cluster to which theclustering center belongs as a re-crime risk prediction model of the test sample, inputting the attribute of the test sample into the re-crime risk prediction model of the test sample, and predictingthe re-crime behavior of the test sample through the model. According to the method, the re-crime prediction effect is more real, effective and accurate, and the calculation speed is higher.

Description

Technical field [0001] The invention relates to the technical field of crime prediction, in particular to a method, medium and computing device for re-crime risk prediction based on reinforcement learning. Background technique [0002] Crime is a social phenomenon in human society. With the continuous progress of human society, especially the rapid development of modern science and technology, crimes have undergone great changes in quantity, scale, crime methods, and the degree of harm to society, posing a threat to human society. It becomes more and more serious. Practice has proved that it is far from enough to crack down on the symptoms of crime. Therefore, people place their hopes on crime prevention. [0003] Parole, temporary execution outside prison, and prison release are three types of special groups who have undergone prison reforms. Due to factors such as poor adaptability to society and unstable psychological state, they are extremely prone to commit crimes again. An...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06Q10/04G06Q10/06G06K9/62G06N3/08G06N3/04
CPCG06Q10/04G06N3/084G06N3/086G06Q10/0635G06N3/045G06F18/23
Inventor 李康顺王梓铭刘嘉豪方鸿铭雷逸舒
Owner SOUTH CHINA AGRI UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products