Loading repository data…
giovannamazzali / repository
Predicting life expectancy across countries using real-world demographic data. This project explores multiple machine learning models and data processing techniques to analyze the impact of different parameters on life expectancy. Built with Python, SQL, and Power BI using data from the World Bank and WHO.
A transparent discovery signal based on current public GitHub metadata.
This score does not audit code, security, maintainers, documentation quality, or suitability. Verify the repository and its current documentation before adoption.
This project investigates how various social, health, environmental, and economic factors influence life expectancy across countries. It leverages multiple machine learning models to predict life expectancy based on real-world data sourced from organizations like the WHO and World Bank.
The main objectives of this project are:
life-expectancy-prediction/
│
├── data/
│ ├── raw/ # Original CSVs from WHO, World Bank, etc.
│ └── processed/ # Cleaned data used for modeling
│
├── dashboard/
│ └── Project.pbix # Final visualization dashboard
│
├── notebooks/
│ ├── export_tables.ipynb # SQL table creation and export logic
│ ├── predictors.ipynb # Model definitions, training, evaluation
│ └── linear_regression_func.ipynb # Linear regression previous problem study
│
├── scripts/
│ ├── config.py # Config file
│ ├── ds_charts.py # Modular functions for charts
│ └── main.py # Helpers
│
└── README.md # This file
The data used for this project comes from:
A total of 40+ CSVs were considered. After evaluating completeness and relevance, 25 datasets were selected and merged. Only columns covering at least 100 countries across 20+ years were included.
Examples of selected features include:
All data was transformed to match a consistent format, with as central reference fields.
Selected from shared topics, language and repository description—not editorial ratings.
reenabikkineni /
Jupyter Notebook for predicting life expectancy using OLS and Logit regression. Includes data preprocessing, EDA, linear and logistic model training, evaluation (RMSE, R², accuracy, ROC), and visualization to analyze factors influencing global life expectancy.
34/100 health| Model | Type | Description |
|---|---|---|
| Linear Regression | Baseline | Simple and interpretable |
| Ridge Regression | Regularized | Handles multicollinearity with L2 penalty |
| Lasso Regression | Regularized | Performs feature selection with L1 penalty |
| Decision Tree | Tree-based | Captures non-linear interactions, prone to overfitting |
| Random Forest | Ensemble | Combines many trees to reduce variance |
| Gradient Boosting | Ensemble | Boosts weak learners sequentially |
| Neural Network | Deep learning | MLP with several hidden layers (Keras) |
After training and evaluating the models, the following patterns emerged:
An interactive Power BI dashboard was created to:
Note: Internal IDs and other authors have been omitted for privacy.
For access to the full technical report, please contact the project owner.
This project is for academic purposes and is not licensed for commercial use.