Topic: Spatiotemporal Data Mining, Representation Learning, Trajectory Data Mining
This project aims to propose pre-training representation learning methods for spatial-temporal trajectory data, modeling multiple features to improve traffic prediction tasks. It demonstrates how trajectory representation learning can enhance traffic data mining.
My major contributions include:
- I conducted pioneering research on pre-training techniques for spatial-temporal data, with relevant results published in TALE and CTLE.
- I originated the concept for this project and wrote the proposal.
- I developed a pre-training representation learning method for trajectory data to improve traffic prediction accuracy, with findings published in MMTEC.
- I independently completed the development of the project’s data mining platform, which serves as a demonstration tool.