Justus-98/kenya-child-health-analysis-

# Spatial-Temporal Analysis of Under-Five Child Health Outcomes in Kenya (2021–2023) ## Overview This project presents a comprehensive data-driven analysis of under-five child health outcomes across Kenya’s 47 counties over a 30-month period (January 2021 – June 2023). The study integrates epidemiological indicators and intervention data to uncover spatial patterns, temporal trends, and underlying drivers of acute malnutrition. The goal is to move beyond descriptive reporting and generate **actionable public health intelligence** for targeted intervention planning. --- ## Key Objectives - Analyze relationships between **diarrhoeal disease, stunting, and acute malnutrition** - Evaluate the effectiveness and targeting of **deworming interventions** - Identify **spatial and temporal patterns** in child health outcomes - Segment counties into **risk profiles using unsupervised machine learning** - Translate statistical findings into **policy-relevant recommendations** --- ## Data Description The dataset includes county-level aggregated indicators: - Acute Malnutrition Cases - Diarrhoea Cases - Total Stunted Children - Total Dewormed Children - Time dimension: Monthly (2021–2023) - Spatial dimension: Kenya (47 counties) --- ## Methodology ### 1. Exploratory & Correlation Analysis - Pearson correlation (linear relationships) - Spearman rank correlation (robust non-parametric validation) ### 2. Inferential Modelling - Multiple Linear Regression: \[ \text{Acute Malnutrition} = \beta_0 + \beta_1(\text{Diarrhoea}) + \beta_2(\text{Stunting}) + \beta_3(\text{Deworming}) + \varepsilon \] - Model diagnostics: - Residual analysis - Normality checks (Q–Q plots) - Heteroskedasticity assessment --- ### 3. Unsupervised Machine Learning - K-Means clustering - Elbow Method (optimal k selection) - Silhouette Analysis (cluster validation) - County-level risk profiling --- ### 4. Spatial & Temporal Analysis - County-level comparative analysis - Temporal trend decomposit …