Egypt Road Accidents Analysis & Forecasting using real CAPMAS data, Excel,Power BI, Python, and time-series models to analyze 2021–2025 trends and forecast 2026.
# Egypt Road Accidents Analysis & Forecasting | 2021–2026
An end-to-end data analysis and forecasting project using real road accident data from Egypt. The project analyzes data from 2021 to 2024, evaluates forecasting models against actual 2025 data, and uses the best-performing model to forecast 2026.
## Project Overview
The project aims to analyze road accident deaths and injuries across Egypt, identify important patterns and affected groups, and provide an interactive Power BI dashboard supported by time-series forecasting.
The workflow covers the full data analysis process, starting from raw government reports and ending with interactive dashboards and a validated forecasting model.
## Data Source
The data was collected from the **Central Agency for Public Mobilization and Statistics (CAPMAS)** in Egypt.
The original data was available in PDF reports, which required extraction, restructuring, cleaning, and transformation before it could be used for analysis.
## Data Preparation
The data preparation process included:
1. Extracting the required tables from CAPMAS PDF reports.
2. Converting the extracted data into Excel.
3. Merging and organizing the required tables.
4. Selecting the relevant tables and fields for analysis.
5. Performing initial data cleaning using Python.
6. Checking for null values and duplicates.
7. Importing the data into Power Query.
8. Unpivoting columns where required.
9. Standardizing data types and formats.
10. Cleaning and standardizing governorate names.
11. Preparing the final dataset for analysis.
## Data Modeling
A relational data model was created in Power BI.
Several dimensions resulted in many-to-many relationships, so bridge tables were created to avoid ambiguous relationships and improve the model structure.
A dedicated Date Table was also created to support time-based analysis and forecasting.
## Power BI Dashboards
The project contains four main dashboard sections:
### Overview
Provides a high-level view of r …