ilyesbhz/Tunisia-Economic-Indicators-Dashboard

# Tunisia Economic Intelligence Dashboard An end-to-end Business Intelligence and Machine Learning project analyzing 60+ years of Tunisia's economic history using World Bank open data. Combines Python ETL, R statistical analysis, scikit-learn forecasting, and Power BI dashboards to tell the story of Tunisia's economic development from independence to 2027. --- ## Key Findings | Indicator | Pre-revolution avg | Post-revolution avg | Change | |---|---|---|---| | GDP Growth | 4.1% | 1.18% | −2.92pp | | Unemployment | 13.4% | 15.9% | +2.5pp | | Inflation | 3.06% | 6.2% | +3.14pp | - **GDP** peaked at $51.3B in 2014, three years after the 2011 revolution, then declined steadily - **Youth unemployment** averages 33.3% — more than double the total rate of 15.2% - **Inflation** hit a record 9.3% in 2023, the worst in modern Tunisian history - **FDI** strongly negatively correlates with unemployment (r = −0.56) — periods of higher foreign investment consistently coincide with lower joblessness - **COVID era** (2020–2024) produced near-zero GDP growth (0.06%) alongside the highest average inflation on record (7.24%) --- ## ML Forecast (2025–2027) Two machine learning models trained on 1986–2024 data forecast a continued slowdown in Tunisia's economy: | Year | GDP Growth (forecast) | Unemployment (forecast) | |---|---|---| | 2025 | 1.40% | 15.35% | | 2026 | 1.14% | 15.57% | | 2027 | 0.95% | 15.74% | **The models project declining GDP growth and slowly rising unemployment through 2027, consistent with the post-revolution structural trend.** ### Model Performance | Model | Algorithm | CV MAE | Train R² | Training samples | |---|---|---|---|---| | Unemployment | Gradient Boosting | 1.56% ± 0.50% | 1.00* | 32 | | GDP Growth | Ridge Regression | 0.72% ± 0.36% | 0.98 | 39 | *High train R² on the unemployment model reflects limited sample size (32 years). The CV MAE of 1.56% is the honest out-of-sample performance estimate. ### Key predictors **Unemployment drivers** ( …