The GIScRG (Geographic Information Science Research Group) and QMRG (Quantitative Methods Research Group) are pleased to announce the winner of our 2017 Exceptional contribution to (GIS/statistical) software competition.

Zhaoya Gong has been awarded £400 for your development of the ARTMAP-based GeoComputation Toolbox.

Further details: The ARTMAP-based GeoComputation Toolbox is a set of ARTMAP-based neural networks tools for spatial data science and geocomputation developed as a QGIS Python Plugin. Recent studies have documented superior performance and accuracy of these models for pattern recognition and soft classi cation of remotely sensed imagery and land-use change prediction [1, 2, 3]. As a workstation version attached to an open source GIS software, this package will make use of common multicore CPUs platforms for parallel computing. Speci cally, this package will be implemented with the support of Cython and OpenMP to generate optimized and parallelized codes that process data in parallel threads with a great computational performance. This type of implementation aims to address the challenges of data and computational intensity for mining large volumes of spatial data (e.g., remote sensing data) with complex machine learning methods.

References
1. Gong, Z., Thill, J.C. and Liu, W., 2015. ART‐P‐MAP Neural Networks Modeling of Land‐Use Change: Accounting for Spatial Heterogeneity and Uncertainty. Geographical Analysis, 47(4), pp.376-409.
2. Liu, W. and Seto, K.C., 2008. Using the ART-MMAP neural network to model and predict urban growth: a spatiotemporal data mining approach. Environment and Planning B: Planning and Design, 35(2), pp.296-317.
3. Liu, W., Seto, K.C., Wu, E.Y., Gopal, S. and Woodcock, C.E., 2004. ART-MMAP: A neural network approach to subpixel classi cation. IEEE transactions on geoscience and remote sensing, 42(9), pp.1976-1983.