# Forecasting Financial Inclusion in Ethiopia
## 📌 Project Overview
This project analyzes and forecasts **financial inclusion trends in Ethiopia**, with a primary focus on **account ownership and access to financial services**. The unified dataset combines observations, policy events, and contextual metadata, aiming to:
- Explore historical trends in financial inclusion
- Contextualize changes using major policy and infrastructure events
- Build a reproducible pipeline for analysis and forecasting
- Transparently document data gaps and limitations
The project emphasizes **analytical correctness and interpretability**, avoiding unsupported causal claims.
---
## 🎯 Objectives
1. **Exploratory Data Analysis (EDA)**
- Analyze trends in account ownership and mobile money access
- Identify data availability and gaps across indicators
- Visualize trends with policy/event overlays
2. **Data Engineering**
- Build a unified, schema-consistent dataset
- Separate observations, events, and (potential) impact links
- Ensure reproducibility and robustness
3. **Forecasting**
- Forecast national account ownership trends
- Use historical access indicators as the primary signal
- Treat policy events as contextual, not causal inputs
4. **Documentation & Transparency**
- Clearly document assumptions and limitations
- Avoid overinterpretation of sparse or missing data
---
## đź“‚ Repository Structure
```plaintext
Forecasting-Financial-Inclusion-in-Ethiopia/
│
├── data/
│ ├── raw/
│ │ ├── ethiopia_fi_unified_data.xlsx
│ │ └── reference_codes.xlsx
│ └── processed/
│
├── notebooks/
│ ├── 01_data_inspection.ipynb
│ ├── 02_exploratory_analysis.ipynb
│ └── 03_forecasting.ipynb
│
├── src/
│ ├── data_loader.py
│ ├── eda.py
│ └── forecasting.py
│
├── figures/
│ └── plots/
│
├── requirements.txt
├── README.md
└── .gitignore
```
---
## đź§ľ Data Description
The dataset follows a unified schema with multiple record types:
### Record Types
- **observation**: Quanti …