Central_Line_delay_system

🚇 Central Line Delay Predictor

A real-time machine learning project that predicts train delays on the London Central Line using live data from the Transport for London (TfL) Unified API

---

📌 Overview

• 🔄 Collects real-time Central Line train data
• 🧹 Cleans and processes the data
• 🧠 Trains a classification model to predict delays (>2 minutes)
• 📈 Visualizes performance and feature importance
• 🌐 Hosts a live dashboard showing the latest predictions

---

🛠 Tech Stack

• Python – Core programming logic
• Pandas, Scikit-learn, RandomForest – Machine learning
• Seaborn, Matplotlib – Data visualization
• Requests – TfL API integration
• Netlify – Frontend deployment
• TfL Unified API – Live train prediction data

---

🔍 Problem Statement

aim to predict whether an incoming Central Line train will be delayed by more than 2 minutes based on:

• Station name
• Time of request
• Platform name
• Scheduled vs actual arrival gap

---

🧠 Prediction Logic

A train is considered delayed if:

expectedArrival - timestamp > 2 minutes