A comparative study of machine learning algorithms for predicting domestic violence vulnerability in Liberian women

BMC Women's Health

Research

Riaz Rahman¹  Zahidul Islam Khan¹  Sabiha Shirin Sara¹  Md. Nafiul Alam Khan¹  Md. Asikur Rahman¹ 

[1] Statistic discipline, Khulna University, 9208, Khulna, Bangladesh;

关键词: XGBoost;    Decision tree;    K-NN;    CatBoost;    Domestic Violence;    Machine learning technique;    Prediction;    Liberia;    DHS;   

DOI  :  10.1186/s12905-023-02701-9

 received in 2023-04-06, accepted in 2023-10-10,  发布年份 2023

来源: Springer

【 摘 要 】

Domestic violence against women is a prevalent in Liberia, with nearly half of women reporting physical violence. However, research on the biosocial factors contributing to this issue remains limited. This study aims to predict women’s vulnerability to domestic violence using a machine learning approach, leveraging data from the Liberian Demographic and Health Survey (LDHS) conducted in 2019–2020.We employed seven machine learning algorithms to achieve this goal, including ANN, KNN, RF, DT, XGBoost, LightGBM, and CatBoost. Our analysis revealed that the LightGBM and RF models achieved the highest accuracy in predicting women’s vulnerability to domestic violence in Liberia, with 81% and 82% accuracy rates, respectively. One of the key features identified across multiple algorithms was the number of people who had experienced emotional violence.These findings offer important insights into the underlying characteristics and risk factors associated with domestic violence against women in Liberia. By utilizing machine learning techniques, we can better predict and understand this complex issue, ultimately contributing to the development of more effective prevention and intervention strategies.

【 授权许可】

CC BY   
© BioMed Central Ltd., part of Springer Nature 2023

【 图 表 】

Fig. 2

Fig. 1

Fig. 1

Fig. 1

12951_2015_155_Article_IEq73.gif

Fig. 3

Fig. 2

Fig. 3

Fig. 1

Fig. 2

Fig. 5