PREDICTION OF ROAD TRAFFIC FROM MULTIPLE SOURCES USING GAUSSIAN APPROACH

Prediction of road traffic from multiple sources using Gaussian approach is most import in intelligent transport systems. Existing works are only focused on non-intrusive sensors that are very expensive. Sensors are detecting traffic conditions and image recognitionetc. The maintains of these sensors are very difficult. To address the issue, this paper aims to improve road traffic speed prediction by using tweet sensors and social media. This includes many challenges, including location uncertainty of low-resolution data, language ambiguity of traffic description in text etc. To response these challenges we provide a uniform modeling probabilistic framework called Topic Enhanced Gaussian Aggregation model (TEGPAM). It consists of three components location disaggregation model, traffic topic model, Traffic speed Gaussian model.