🔌 Electrical Grid Stability Predictor

A machine learning project using regression and classification models to analyze and predict the stability of an electrical grid system.

📄 Project Description

This repository contains a comprehensive machine learning analysis of the Electrical Grid Stability Simulated Data from the UCI Machine Learning Repository.

The project explores both regression and classification approaches to model grid stability:

• ✅ 5 Regression Models
• ✅ 5 Classification Models
• ✅ Hyperparameter Tuning on top 3 models in each category
• ✅ Model Evaluation using metrics and visualizations
• ✅ Comparative Analysis of all models

🔍 Objectives:

• Predict the stability score (stab - a continuous variable)
• Classify the system as stable or unstable (stabf - categorical)

---

📊 Dataset Overview

• Source: UCI Electrical Grid Stability Simulated Data
• Features: 12 predictors including:
  • tau1–tau4: Reaction time constants
  • p1–p4: Nominal power values
  • g1–g4: Price elasticity coefficients
• Targets:
  • stab: Real-valued indicator of grid stability
  • stabf: Categorical label (stable or unstable)

---

📈 Project Highlights

• Performance metrics: R², MSE, MAE for regression; Accuracy, F1-score, and confusion matrix for classification
• Model visualizations and comparison plots
• Clean workflow and well-commented code in Colab

---

🚀 Future Ideas

• Implement feature selection techniques
• Add a web interface with Streamlit
• Explore deep learning models (MLPs)

---