UCL logo

UCL Discovery

UCL home » Library Services » Electronic resources » UCL Discovery

Accommodating spatial associations in DRNN for space-time analysis

Cheng, T; Wang, JQ; (2009) Accommodating spatial associations in DRNN for space-time analysis. COMPUT ENVIRON URBAN , 33 (6) 409 - 418. 10.1016/j.compenvurbsys.2009.08.004.

Full text not available from this repository.


Dynamic recurrent neural networks (DRNN) are neural networks with feedback connections. They are superior to static feedforward neural networks (SFNN) in nonlinear time-series analysis because they can also accommodate temporal associations. However, like SFNN, DRNN presents a black box approach to space-time analysis. This paper seeks to incorporate spatial associations into a DRNN, through its structure and initial weights. It suggests a novel approach to defining the topological structure and initial weights of DRNN based on the spatial associations of spatial units. This is seen as vital for improving the accuracy and efficiency of prediction and forecasting using space-time models. The proposed method is illustrated using three instances of space-time analysis, which are each characterized by different spatial data types (discrete and continuous). Computational accuracy and efficiency are much improved by incorporating spatial associations in DRNN. This reveals that DRNN can be a powerful tool for modeling space-time series with complex spatial and temporal characteristics. (C) 2009 Elsevier Ltd. All rights reserved.

Type: Article
Title: Accommodating spatial associations in DRNN for space-time analysis
DOI: 10.1016/j.compenvurbsys.2009.08.004
Keywords: Spatial associations, DRNN, Space-time analysis, Spatial weight matrix, NEURAL-NETWORKS, SYSTEMS
URI: http://discovery.ucl.ac.uk/id/eprint/76089
Downloads since deposit
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item