No description available

🚨 AI-powered traffic accident severity predictor using Random Forest ML. Built with Python, Streamlit & Plotly. Features real-time prediction, interactive dashboards, risk radar charts & 22-feature weather/road analysis.

“Unsupervised ML project to cluster drivers and vehicles based on accident risk patterns using RTA dataset.”

Geospatial ML pipeline from my MSc dissertation: risk-mapping for orangutan poaching using forest-loss, access proxies (roads/settlements), and conservation context (protected areas).

React + Mapbox app to assess route accident risk. Pick start/end; backend derives curvature/lanes/road type, lighting (sunrise/sunset) and weather (NWS), runs an ML model, and returns risk. Terraform config included for AWS deployment.

Predictive Analysis of Road Collisions in Halifax (HRM) This project uses ML and explainable AI techniques to predict the severity of road collisions. It tackles class imbalance with SMOTE/ADASYN and compares multiple classifiers. The goal is to identify high-risk patterns and support data-driven road safety planning.

AI-Driven Road Accident Prediction System (Mar 2025): Developed a hybrid rule-based and ML (Random Forest) system with 85% accuracy, integrated 3 live APIs for real-time risk analysis, enabled hotspot visualization via geospatial heatmaps, and improved model accuracy by 15% through real-world feedback loops.

An intelligent system that uses AI/ML 🤖🧠 to analyze traffic conditions 🚦🛣️, driver stress 😰📈, and behavioral patterns 👀📊 to predict accident risk ⚠️🚨 in real time ⏱️. Designed to enhance driver safety 🛡️🚗, reduce road accidents 🚑💥, and support smarter driving decisions 📊✅.

No description available

ML Project (Road Accident Risk Predicition)

A ML model estimating risky roads implemented as a final project for ID2223 at KTH

No description available

No description available

Проект по прогнозированию дорожных рисков, ML-модель предсказывает вероятность ДТП с точностью 88.5%

Machine learning system for predicting high-risk road segments using geospatial data, feature engineering, and ensemble models (RF, XGBoost, Logistic Regression)

Using combination of models to predict road accident risk

Road Accident Risk Analysis using ML and Hotspot Clustering

A predictive machine learning framework for classifying road accident risk using traffic, weather, and historical crash data.

Interactive 2D road network risk analysis platform with ML-powered predictions

ML-based accident risk prediction using weather, traffic, and road features

ML pipeline to identify road accident risk factors in autonomous vehicle contexts

Road Accident Prediction & Risk Analysis using ML. Analyze historical data, predict accident severity, and gain insights to improve road safety and reduce risks.

This project predicts road accident risk using ML and ANN models. It analyzes features like road type, weather, lighting, and speed limit to identify accident-prone conditions. Models such as Random Forest, CatBoost, and ANN were compared, achieving up to 88.5% accuracy and 0.0589 RMSE, ensuring reliable

No description available

Predictive ML solution (XGBoost + feature engineering + hyperparameter tuning) for road-segment accident risk estimation

Kaggle project for Road Risk ML predictions (https://www.kaggle.com/competitions/playground-series-s5e10/data)

End-to-end ML pipeline for road accident risk prediction with synthetic data and Bayesian correction.

This research presents an NLP/ML system for automated identification of contractual risks and recommendation of risk responses in road infrastructure projects.

Interactive Streamlit app to compare two roads and predict accident risk based on road, traffic, and environmental features. Includes ML model predictions or a fallback formula for realistic risk scoring.

ADAS ML platform for road scene risk classification with training, evaluation, FastAPI inference, and failure analysis dashboard