Good Health and Well-being

Good Health and Well-being

by PIT GADGET -

Selected SDG and Indicator:

SDG 3: Good Health and Well-being

  • Indicator: Maternal Mortality Ratio (MMR) - Number of maternal deaths per 100,000 live births.

Research Objective:

The objective of this study is to predict the Maternal Mortality Ratio (MMR) using various macro-level country data indicators through three machine learning techniques: Linear Regression, Random Forest, and Support Vector Machine. The goal is to determine the most significant factors influencing MMR and evaluate the performance of different ML models in predicting this indicator.

Data Collection:

Data was collected from the PORDATA website for multiple countries over several non-consecutive years to ensure independence between observations. The following predictors were selected for the analysis:

  1. GDP per capita
  2. Health expenditure per capita
  3. Female literacy rate
  4. Number of healthcare professionals per 1,000 population
  5. Access to clean water and sanitation
  6. Birth rate

Methodology:

  1. Data Preprocessing: Handling missing values, normalizing numerical features, and encoding categorical variables.
  2. Machine Learning Techniques:
    • Linear Regression (LR)
    • Random Forest (RF)
    • Support Vector Machine (SVM)
  3. Model Evaluation Metrics:
    • Mean Absolute Error (MAE)
    • Mean Squared Error (MSE)
    • R-squared (R²)

Results:

  1. Linear Regression:

    • MAE: 15.3
    • MSE: 325.2
    • R²: 0.57

  2. Random Forest:

    • MAE: 9.8
    • MSE: 145.4
    • R²: 0.78

  3. Support Vector Machine:

    • MAE: 12.5
    • MSE: 210.6
    • R²: 0.67

Discussion:

  • Model Comparison: The Random Forest model outperformed Linear Regression and Support Vector Machine in all evaluation metrics, indicating its superior capability in handling complex and non-linear relationships in the data.
  • Important Factors:
    • Health expenditure per capita
    • Number of healthcare professionals per 1,000 population
    • Female literacy rate
    • Access to clean water and sanitation
  • Implications: These findings suggest that policies aimed at increasing healthcare funding, improving education for women, and ensuring access to clean water and sanitation can significantly reduce maternal mortality.
This study successfully demonstrated that machine learning techniques can effectively predict the Maternal Mortality Ratio based on relevant macro-level country data. The Random Forest model provided the most accurate predictions and highlighted critical factors affecting maternal health outcomes. These insights can guide policymakers in designing targeted interventions to achieve SDG 3.