District-Level Shadow Economy Risk Assessment using NASA VIIRS Nighttime Lights • RBI Financial Infrastructure • Census of India • MSME Data • Explainable Machine Learning
The informal (shadow) economy contributes significantly to India's economic activity but remains difficult to quantify due to its unreported nature.
SCIEQS (Satellite-Calibrated Informal Economy Quantification System) is a data analytics and explainable machine learning project that estimates district-level informal economic activity by integrating satellite-derived nighttime light intensity with financial, demographic, and industrial datasets.
Instead of relying on traditional surveys, SCIEQS combines multiple proxy indicators into a statistically validated composite index capable of identifying economically active yet financially underrepresented regions across India.
✔ Quantify district-level shadow economy intensity ✔ Detect economically active but financially underrepresented regions ✔ Build an explainable Composite Shadow Economy Index ✔ Apply dimensionality reduction using Principal Component Analysis (PCA) ✔ Discover economic patterns through K-Means Clustering ✔ Deliver executive dashboards for policy-level decision making
- District Shadow Economy Risk Map
- Top High-Risk District Rankings
- State-wise Risk Comparison
- KPI Cards
- Business Insights
- PCA Feature Importance
- Statistical Validation
- Cluster Distribution
- Economic Formalization Gap
- Machine Learning Interpretation
Raw Datasets
│
▼
Data Cleaning
│
▼
Feature Engineering
│
▼
Normalization
│
▼
Composite Shadow Index
│
▼
Principal Component Analysis
│
▼
K-Means Clustering
│
▼
Business Intelligence Dashboard
| Dataset | Purpose |
|---|---|
| 🛰 NASA VIIRS Nighttime Lights | Economic Activity Proxy |
| 🏦 RBI Financial Infrastructure | Financial Inclusion |
| 👥 Census of India | Population Statistics |
| 🏭 MSME Database | Industrial Activity |
| ⚡ Electricity Consumption | Economic Development Proxy |
✔ Feature Engineering → Normalization → Composite Index Construction → PCA → K-Means Clustering → Cluster Interpretation → Statistical Validation
- Principal Component Analysis (PCA)
- Min-Max Scaling
- Composite Weighted Index
- K-Means Clustering
- Correlation Analysis
- Distribution Analysis
- Feature Importance / Explainable Analytics
| Category | Technologies |
|---|---|
| Programming | Python |
| Data Processing | Pandas, NumPy |
| Machine Learning | Scikit-learn |
| Visualization | Tableau |
| Notebook | Jupyter |
| Version Control | Git & GitHub |
SCIEQS/
│
├── assets/ # Dashboard screenshots used in this README
├── data/ # Raw and processed datasets
├── notebooks/ # Jupyter analysis notebook(s)
├── outputs/ # Generated charts, exports, results
├── tableau/ # Tableau extract + packaged workbook (.twbx)
├── docs/ # Any supporting documentation
├── requirements.txt
├── LICENSE
└── README.md
git clone https://github.com/neeldas0032/SCIEQS.git
cd SCIEQS
pip install -r requirements.txt
jupyter notebook👉 Tableau Public — Executive Dashboard
📌 Nighttime light intensity strongly correlates with regional economic activity. 📌 Financial infrastructure alone cannot explain economic formalization. 📌 PCA reduced feature redundancy while preserving most information. 📌 Four economically distinct regional clusters emerged through K-Means clustering. 📌 High-risk districts exhibit strong economic activity but comparatively weak financial inclusion.
SCIEQS demonstrates how satellite remote sensing and socioeconomic indicators can be integrated into an explainable analytics framework for:
- Public Policy
- Financial Inclusion
- Regional Development
- Economic Intelligence
- Urban Planning
- Development Economics
- XGBoost-based Risk Prediction
- Temporal Nightlight Analysis
- SHAP Explainability
- Streamlit Web Application
- Real-time Dashboard
- State-Level Forecasting
B.Tech Computer Science & Engineering (AI & ML) Data Analytics • Machine Learning • Business Intelligence • Geospatial Analytics
This project is licensed under the MIT License — see the LICENSE file for details.