Linear regression analysis of Kenya youth unemployment (1991–2030)
# Kenya Unemployment Analysis (World Bank Data)
## Project Overview
This project explores unemployment trends in Kenya using World Bank data.
It involves cleaning raw CSV data, visualizing year-by-year unemployment changes, and using a Linear Regression model to forecast future rates up to 2030.
### Objectives
Understand the historical trend of unemployment in Kenya
Visualize year-over-year changes in unemployment rates
Build a predictive model to estimate unemployment through 2030
Derive insights useful for policymakers and development organizations
### Tools & Libraries
`Python`
`Pandas` — data cleaning and transformation
`Matplotlib` — visualization
`Scikit-learn` — linear regression modelling
`NumPy` — numerical computations
### Dataset
Source: World Bank Open Data
Indicator: Unemployment, total (% of total labour force)
Time Range: 1991–2023
Data Format: CSV
### Steps & Process
#### Data Loading & Cleaning
Read World Bank CSV, skip metadata rows
Filter for Kenya only
Drop irrelevant columns (Country Code, Indicator Code, etc.)
Handle missing and blank values
#### Exploratory Analysis
Visualized Kenya’s unemployment trends over time
Identified years with spikes or dips
#### Modeling
Trained a Linear Regression model on unemployment vs year
Evaluated model coefficients and fitted line
Predicted unemployment up to 2030
#### Visualization
Plotted actual vs predicted unemployment trends
Added a dashed line for 2025–2030 projections
### Key Insights
Kenya’s unemployment rate has shown fluctuations over time with a general trend visible in the regression line.
The model suggests a moderate upward/downward trend (update this based on your actual results).
Forecasts highlight possible patterns that can inform youth and labor market planning.
### Folder Layout
- **Kenya_unemployment.ipynb** → main notebook for analysis
- **unemployment.csv** → cleaned dataset
- **README.md** → project overview and documentation
- **requirements.txt** …