Home

Pre-term Birth Prediction – Sri Lankan Population

Welcome to the GitHub Wiki for the Preterm Birth Prediction project. This work aims to predict the likelihood of pre-term birth using a comprehensive dataset collected from Sri Lankan hospitals.

Overview

Pre-term birth (PTB), defined as birth before 37 weeks of gestation, is a leading cause of neonatal mortality and long-term complications. This project focuses on developing a machine learning model tailored specifically for the Sri Lankan population, using locally gathered clinical and demographic data.

Dataset Description

• Source: Data collected from multiple Sri Lankan hospitals.
• Total Features: 137 features.
• Time Frame: Data spans from the 12th week of gestational age through to neonatal outcomes.
• Data Type: Mixed types – includes clinical, behavioral, demographic, and outcome-related features.

Objectives

• Build a reliable model to predict the risk of preterm birth as early as possible during pregnancy.
• Address challenges specific to Sri Lankan maternal health data, such as:

1. Class imbalance
2. Limited and localized dataset
3. Diverse socio-demographic profiles

Methodology (To Be Expanded)

• Data Preprocessing
• Feature Selection and Engineering
• Model Training and Evaluation
• Handling Imbalanced Data
• Performance Metrics and Validation

Impact

• Assist healthcare providers in early risk identification.
• Contribute towards personalized maternal care.
• Enable more data-driven interventions in Sri Lanka's maternal health system.