Understanding travel behaviour and demand is of constant importance to transportation communities agencies in every country. Nowadays, attempts have been made automatically infer modes from positional data, such as the data collected by using GPS devices so that cost time budget conventional diary survey could be significantly reduced. Some limitations, however, exist literature, aspects collection (sample size selected, duration study, granularity data), selection variables (or combination...

As more and real time spatio-temporal datasets become available at increasing spatial temporal resolutions, the provision of high quality, predictive information about processes becomes an increasingly feasible goal. However, many sensor networks that collect are prone to failure, resulting in missing data. To complicate matters, data is often not random, characterised by long periods where no observed. The performance traditional univariate forecasting methods such as ARIMA models decreases...

Various statistical model specifications for describing spatiotemporal processes have been proposed over the years, including space–time autoregressive integrated moving average ( STARIMA ) and its various extensions. These assume that correlation in data can be adequately described by parameters are globally fixed spatially and/or temporally. They inadequate cases which correlations among dynamic heterogeneous, such as network data. The aim of this article is to describe autocorrelation...

Accurate and reliable forecasting of traffic variables is one the primary functions Intelligent Transportation Systems. Reliable systems that are able to forecast conditions accurately, multiple time steps into future, required for advanced traveller information systems. However, a difficult task because nonlinear nonstationary properties series. Traditional linear models incapable modelling such properties, typically perform poorly, particularly when differ from norm. Machine learning...

Extracting information related to weather and visual conditions at a given time space is indispensable for scene awareness, which strongly impacts our behaviours, from simply walking in city riding bike, driving car, or autonomous drive-assistance. Despite the significance of this subject, it has still not been fully addressed by machine intelligence relying on deep learning computer vision detect multi-labels with unified method that can be easily used practice. What achieved to-date are...

Modelling the spread of coronavirus globally while learning trends at global and country levels remains crucial for tackling pandemic. We introduce a novel variational-LSTM Autoencoder model to predict each across globe. This deep Spatio-temporal does not only rely on historical data virus but also includes factors related urban characteristics represented in locational demographic (such as population density, population, fertility rate), an index that represents governmental measures...

Cities are complex systems that constantly changing. This paper explores the capabilities of using crowdsourced street-level imagery in observing city dynamics. Visual walkability is an example such index, where different results may be obtained depending on locational and temporal factors. introduces a new index called Type Walkability (TVW) to characterize classify visual Inner London utilizing Mapillary images. The method based panoptic segmentation, pixel-level segmentation instance...

Species abundance regression models have been fitted with the GLIM computer package. These take into account discrete nature of dependent variable, numbers species. The inappropriateness conven- tional normal is demonstrated by use a probability plot.

Visualisation is an effective tool for studying traffic congestion using massive datasets collected from sensors. Existing techniques can reveal where/when congested areas are formed, developed, and moved on one or several highway roads, but it still challenging to visualise the evolution of whole road network, especially dense urban networks. To address this challenge, paper proposes three 3D exploratory visualisation techniques: isosurface, constrained wall map. These have different...

Within the burgeoning expansion of deep learning and computer vision across different fields science, when it comes to urban development, applications are still limited towards notions smart cities autonomous vehicles. Indeed, a wide gap knowledge appears regions in less developed countries where chaos informality is dominant scheme. How can Artificial Intelligence (AI) untangle complexities advance modelling our understanding cities? Various questions debates be raised concerning future...

Model specification is a crucial factor in regression. We show that log-linear models, although the estimated relationship represents conditional expectation of ln Y, antilogarithm does not give Y. This has important implications spatial analysis.

The coronavirus pandemic that started in 2019 has had wide-ranging impacts on many aspects of people's daily lives. At the peak outbreak, lockdown measures and social distancing changed ways which cities function. In particular, they profound urban transportation systems, with public transport being shut down cities. Bike share systems (BSS) were widely reported as having experienced an increase demand during early stages before returning to pre-pandemic levels. However, studies published...

Abstract Modelling the spread of coronavirus globally while learning trends at global and country levels remains crucial for tackling pandemic. We introduce a novel variational LSTM-Autoencoder model to predict each across globe. This deep spatio-temporal does not only rely on historical data virus but also includes factors related urban characteristics represented in locational demographic (such as population density, population, fertility rate), an index that represent governmental...