Urban Growth Modelling with Artificial Neural Network and Logistic Regression. Case Study: Sanandaj City, Iran

Journal Title: Romanian Review of Regional Studies - Year 2013, Vol 9, Issue 2

Abstract

Cities have shown remarkable growth due to attraction, economic, social and facilities centralization in the past few decades. Population and urban expansion especially in developing countries, led to lack of resources, land use change from appropriate agricultural land to urban land use and marginalization. Under these circumstances, land use activity is a major issue and challenge for town and country planners. Different approaches have been attempted in urban expansion modelling. Artificial Neural network (ANN) models are among knowledge-based models which have been used for urban growth modelling. ANNs are powerful tools that use a machine learning approach to quantify and model complex behaviour and patterns. In this research, ANN and logistic regression have been employed for interpreting urban growth modelling. Our case study is Sanandaj city and we used Landsat TM and ETM+ imageries acquired at 2000 and 2006. The dataset used includes distance to main roads, distance to the residence region, elevation, slope, and distance to green space. Percent Area Match (PAM) obtained from modelling of these changes with ANN is equal to 90.47% and the accuracy achieved for urban growth modelling with Logistic Regression (LR) is equal to 88.91%. Percent Correct Match (PCM) and Figure of Merit for ANN method were 91.33% and 59.07% and then for LR were 90.84% and 57.07%, respectively.

Authors and Affiliations

SASSAN MOHAMMADY, MAHMOUD DELAVAR, BRYAN PIJANOWSKI

Keywords

Related Articles

The Demographic Situation in the Northern Economic Region of the Republic of Moldova

The demographic situation within the Northern Economic Region is worse than the national average. This fact is determined by the decrease of the birth rate values, the increase of the mortality values and the negative na...

The Borderlessness of Economic Life and Intended Regionalisation. Thoughts about Boundaries and Regions

The long area from the Baltic to the Adriatic and the Black Sea between the German and the Russian (and the 19th century Turkish) empires is called Eastern-Central Europe in the Hungarian academic discourse. The peoples...

Accessibility to Nodes of Interest: A Practical Application of the Various Forms of the Impedance Curves

Impedance is, in territorial statistics, the greater or lesser difficulty of a territory to be served/provided with adequate infrastructure and related services, useful for the improvement of the resident’s quality of li...

The Impact of Changing Land Use upon the Environment in the Metropolitan Area of Bucharest. Preliminary Considerations

In this paper we highlight the major changes in land use during the transition from the centralised to the market economy (1989-2005). Agricultural de-collectivisation and privatisation have caused major structural chang...

Inter-Regional Disparities in the European Union

One of the European Union's major goals is to reduce disparities between countries and regions. We compute sigma-convergence at national and regional level (NUTS2) on GDP per capita and employment rates to analyze the ev...

Download PDF file
  • EP ID EP109855
  • DOI -
  • Views 160
  • Downloads 0

How To Cite

SASSAN MOHAMMADY, MAHMOUD DELAVAR, BRYAN PIJANOWSKI (2013). Urban Growth Modelling with Artificial Neural Network and Logistic Regression. Case Study: Sanandaj City, Iran. Romanian Review of Regional Studies, 9(2), 47-60. https://europub.co.uk/articles/-A-109855