jm-wanja/outage-prediction-and-grid-reliability

Machine learning project for predicting power outages using KPLC data from Kenya # Outage Prediction and Grid Reliability for Kenya Power ## 🎯 Project Overview This project uses historical outage patterns from Kenya Power and Lighting Company (KPLC) to predict system-wide failures and quantify large-scale outage probabilities. The goal is to deliver actionable insights to strengthen grid resilience and inform preventive actions. ### 🔍 Key Features - **Spatial-Temporal Analysis**: Leverage geographic coordinates and timestamps from historical outage data - **Predictive Modeling**: Machine learning models to forecast outage probabilities - **Interactive Visualization**: Risk maps and dashboards for stakeholders - **Reliability Metrics**: System-wide outage probability, regional risk assessment, MTBO analysis ## 📊 Dataset The project uses the KPLC Electricity Interruption Data from Kaggle, containing: - Geographic coordinates (latitude, longitude) - Timestamp data (ISO dates and Unix timestamps) - Historical outage patterns across Kenya's power grid ## 🛠️ Technology Stack - **Python 3.9+** - Core programming language - **Pandas & NumPy** - Data manipulation and analysis - **Scikit-learn** - Machine learning algorithms - **XGBoost/LightGBM** - Advanced ML models - **GeoPandas & Folium** - Geospatial analysis and mapping - **Matplotlib & Plotly** - Data visualization - **Streamlit** - Web application framework - **FastAPI** - API development ## 🚀 Quick Start ### Prerequisites - Python 3.9 or higher - Git - (Optional) Conda for environment management ### Installation 1. **Clone the repository** ```bash git clone github.com cd outage-prediction-grid-reliability ``` 2. **Set up the environment** Using conda (recommended): ```bash conda env create -f environment.yml conda activate outage-prediction ``` Or using pip: ```bash python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate pip install -r requirements.txt ``` 3. **Install the package in devel …