Predicting mothers' exclusive breastfeeding for the first 6 months: Interface creation study using machine learning...

Area of Science:

• Public Health
• Maternal and Child Health
• Health Informatics

Background:

• Machine learning techniques (MLT) are powerful tools for analyzing large datasets to identify complex patterns.
• Predictive modeling in maternal health can support evidence-based interventions.

Purpose of the Study:

• To develop a machine learning-based prediction model for identifying mothers exclusively breastfeeding for the first six months.
• To create a user-friendly interface for this predictive tool.

Main Methods:

• A dataset of 514 mothers with infants aged 6-24 months was utilized.
• Data from 70% of participants were used to train the machine learning model, with the remaining 30% used for testing.
• The Random Forest Classifier algorithm was selected as the optimal model.

Main Results:

• The Random Forest Classifier demonstrated effectiveness in predicting exclusive breastfeeding.
• Top predictive factors included: maternal health during pregnancy, social support, daily water intake, perceived milk sufficiency, and ease of infant breastfeeding.
• The model identified key variables influencing the likelihood of exclusive breastfeeding.

Conclusions:

• The developed prediction model can facilitate early identification of mothers at risk of not exclusively breastfeeding.
• This early detection allows for timely monitoring and targeted support for mothers in high-risk groups.
• Implementing this MLT tool can enhance public health strategies for promoting breastfeeding.