Related Experiment Video
Updated: Jan 10, 2026
07:31
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
7.5K
Predicting and explaining recurrent child abuse using interpretable machine learning: Evidence from national-level
Donghun Kim1, Ting Jiang2, Kihyun Kim3
1School of Information Management, Nanjing University, No.163, Xianlin Road, Qixia District, Nanjing, China.
Social Science & Medicine (1982)
|November 23, 2025
Summary
Machine learning models predict child abuse recurrence risk. Key factors identified include perpetrator age and abuse type, aiding targeted interventions for child protection.
Area of Science:
- Child protection research
- Machine learning applications in social science
- Public health surveillance
Background:
- Child abuse recurrence poses significant risks to child welfare.
- Understanding predictors of recurrent abuse is crucial for effective intervention.
- National data provides a robust basis for analyzing child abuse patterns.
Purpose of the Study:
- To develop machine learning models for predicting recurrent child abuse risk.
- To identify key risk factors specific to each abuse type (physical, emotional, sexual, neglect).
- To explain individual recurrence risk and inform tailored prevention strategies.
Main Methods:
- Utilized national child abuse data encompassing children, perpetrators, and reported cases.
- Developed distinct predictive models for recurrent physical abuse (RPA), recurrent emotional abuse (REA), recurrent sexual abuse (RSA), and recurrent neglect (RN).
- Analyzed a wide array of factors including child demographics, perpetrator characteristics, abuse episode details, and service interventions.
Main Results:
- Achieved highest predictive accuracy for recurrent neglect (AUC-ROC 0.793), followed by RSA (0.749), REA (0.702), and RPA (0.700).
- Younger perpetrator age was associated with increased risk across all abuse types.
- Counseling for children and perpetrators mitigated risks for RPA, REA, and RSA.
- Same-type abuse recurrence was highly influential for RPA, REA, and RN; perpetrator gender (male) was significant for RSA.
- Parenting attitudes, knowledge, skills, and family environment conflicts were critical for RSA.
Conclusions:
- Machine learning models offer valuable tools for predicting child abuse recurrence.
- Identifying specific risk factors allows for tailored prevention and intervention strategies.
- Findings support child protection agencies in early detection and customized support for at-risk families.
Related Concept Videos
01:18Steps in Outbreak Investigation
472
In the ever-evolving field of public health, statistical analysis serves as a cornerstone for understanding and managing disease outbreaks. By leveraging various statistical tools, health professionals can predict potential outbreaks, analyze ongoing situations, and devise effective responses to mitigate impact. For that to happen, there are a few possible stages of the analysis:
472
01:11Residuals and Least-Squares Property
8.9K
The vertical distance between the actual value of y and the estimated value of y. In other words, it measures the vertical distance between the actual data point and the predicted point on the line
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
8.9K