Traffic Predictive Analysis Through Data Stream Mining
暂无分享,去创建一个
With a huge increase in computational power, Traffic Predictive Analysis has seen various improvements in the recent years. Additionally, this field is experimenting an increase in available data, which allows to produce more precise forecasting and classification models. However, this means that the available data has also seen a huge increase in terms of storage size. Data Stream Mining provides a brand new approach to data processing, allowing to create adaptive, incremental models that do not need huge amounts of storage size, as the data is processed as it is received. In this communication, we will explore the state of the art and the first research efforts that can be found in this direction.
[1] Byoung-Jo Yoon,et al. High-Speed Data-Driven Methodology for Real-Time Traffic Flow Predictions: Practical Applications of ITS , 2018 .
[2] Albert Bifet,et al. Adaptive learning and mining for data streams and frequent patterns , 2009, SKDD.
[3] Pallaviram Sure,et al. Low-Dimensional Models for Traffic Data Processing Using Graph Fourier Transform , 2018, Computing in Science & Engineering.
[4] João Gama,et al. A survey on concept drift adaptation , 2014, ACM Comput. Surv..