Predicción del tiempo de llegada de vehículos basada en aprendizaje automático

#IMTResearchSeminars | Presented by Alejandro Ascencio Laguna, Head of the Intelligent Transportation Systems Unit of the Integrated Transportation and Logistics Coordination at IMT. Machine learning (ML) and geospatial clustering have traditionally been used as independent approaches to address urban freight transportation problems, particularly in predicting arrival times under just-in-time logistics schemes. However, the integration of both approaches has been little explored, while traditional methods based solely on distance metrics offer insufficient accuracy for operational logistics applications. This study proposes a hierarchical framework that combines geographic clustering using k-means as a spatial segmentation mechanism, along with a Random Forest model enhanced through time-enhancing feature engineering. The developed architecture is computationally efficient and robust against the uncertainty inherent in real-world environments. ⏳ Don't miss the opportunity to stay up-to-date! 📚 Subscribe to our channel for more important information! 📺 And of course, join us on LinkedIn, Facebook, and Twitter. 🌐 Expand your knowledge in this exciting field. 💡