manzidelphin/National-Health-Insuarance-Rwanda-Manzi-Delphin-26021-

# 🏥 National Health Insurance – Big Data Capstone Project **Name:** Manzi Delphin **ID:** 26021 **Group:** E **Course:** Introduction to Big Data Analytics **Instructor:** Eric Maniraguha **Academic Year:** 2024–2025, Semester III --- ## 📁 Project Files & Links - 🔗 **Dataset Used**: Health Insurance Dataset (CSV) - 📊 **Power BI Dashboard**: Power BI Project Folder - 📽️ **PowerPoint Presentation**: Capstone Slides --- ## ✅ PART 1: PROBLEM DEFINITION & PLANNING ### 🏷️ I. Sector Selection ☑ **Health** ### ❓ II. Problem Statement > This project investigates disparities and patterns in national health insurance coverage across countries using survey-based health indicators. Using big data analytics, it aims to identify trends across demographics, regions, and years to support equitable policy decision-making. ### 📦 III. Dataset Identification - **Dataset Title:** National Health Insurance – Global DHS Indicators - **Source Link:** Dataset Download - **Number of Rows and Columns:** 15,000+ rows × 20 columns - **Data Structure:** ☑ Structured (CSV) - **Data Status:** ☑ Requires Preprocessing --- ## 🐍 PART 2: PYTHON ANALYTICS TASKS ### ➡️ 1. Clean the Dataset We began by inspecting for missing values, inconsistent formats, and extreme outliers. String values were normalized, and outliers in numerical fields were capped using IQR. ### ➡️ 2. Apply Data Transformations Categorical values were encoded using LabelEncoder, and numerical features such as age, household size, and insurance value were scaled using StandardScaler to ensure modeling fairness. --- ### ➡️ 3. Conduct Exploratory Data Analysis (EDA) We analyzed central tendencies, distributions, and correlations between variables. This helped us detect relationships and understand variable behavior. #### Descriptive Statistics #### Distribution Visuals We used histograms and boxplots to observe feature spread and detect skews and group variances. #### Correlation Analysis This heatmap revea …