Using Predictive ML models to predict Covid trends in Africa
**COVID-19 Spread in Africa Using Machine Learning**
This project analyzes COVID-19 cases and deaths across African countries from 2020 to 2022. It uses exploratory data analysis, K-Means clustering, PCA, Decision Tree Regression, and Random Forest Regression to study regional patterns and predict COVID-19 mortality.
The dataset was obtained from Our World in Data and filtered to include African countries.
**Tools Used**
• Python
• Jupyter Notebook
• Pandas
• NumPy
• Matplotlib
• Seaborn
• Scikit-learn
**Project Steps**
1. Cleaned and filtered the dataset
2. Compared COVID-19 cases and deaths across African regions
3. Used K-Means clustering to group similar countries
4. Used PCA to visualize the clusters
5. Built Decision Tree and Random Forest regression models
6. Evaluated the models using MSE, RMSE, and R²
**Results**
Model RMSE R2 Score
Decision Tree Regressor 1.1321 0.67849
Random Forest Regressor 1.3257 0.5524
The Decision Tree Regressor performed better and explained about 67% of the variation in COVID-19 deaths.
The clustering results also showed that countries with similar COVID-19 outcomes were not always located in the same region.