This article looks at two elements of geographic analysis applied to urban processes – GISystems, software for analyzing geographic information, and GIScience, theories and methods to interpret geographic information. Self-Organizing Maps (SOM) are a type of Artificial Neural Network (ANN), initially studied by Kohonen, Hynninen, Kangas & Laaksonen (1995), capable of automatically organizing data within an artificial space in such a way as to group together data with similar characteristics while distancing dissimilar data. They are therefore able to preserve the topological relationships of archived data and produce output maps that can be analyzed using a GISystem. Because the “Solutions for environmental contrasts in coastal areas. Global Change, human mobility and sustainable urban development” project (SECOA) used data originating from the experiences of the natural and social sciences, the data collection and processing stages were complex. Complexity is characteristic of our society and, to explain phenomena related to us, it is increasingly necessary to analyze all the available data, whatever their origin: if these phenomena exist, they are relevant in one way or another to the formation of reality. So they must be identified and measured using the expedients that make it possible to subsequently include them in the rest of the information. SECOA considered two levels of phenomena: global ones such as climate change and human mobility, and local ones that emerge in the form of conflicts, the dynamics of urban growth, and natural, environmental, social and economic characteristics at the local level. The local is reliant on the global in that it is the positive or negative result of decisions taken and policies applied at a higher level. But the local level also concentrates all the consequences of improper management at the global level while contributing, through its errors and positive behaviors, to changes at the global level.
This article looks at two elements of geographic analysis applied to urban processes – GISystems, software for analyzing geographic information, and GIScience, theories and methods to interpret geographic information. Self-Organizing Maps (SOM) are a type of Artificial Neural Network (ANN), initially studied by Kohonen, Hynninen, Kangas & Laaksonen (1995), capable of automatically organizing data within an artificial space in such a way as to group together data with similar characteristics while distancing dissimilar data. They are therefore able to preserve the topological relationships of archived data and produce output maps that can be analyzed using a GISystem. Because the “Solutions for environmental contrasts in coastal areas. Global Change, human mobility and sustainable urban development” project (SECOA) used data originating from the experiences of the natural and social sciences, the data collection and processing stages were complex. Complexity is characteristic of our society and, to explain phenomena related to us, it is increasingly necessary to analyze all the available data, whatever their origin: if these phenomena exist, they are relevant in one way or another to the formation of reality. So they must be identified and measured using the expedients that make it possible to subsequently include them in the rest of the information. SECOA considered two levels of phenomena: global ones such as climate change and human mobility, and local ones that emerge in the form of conflicts, the dynamics of urban growth, and natural, environmental, social and economic characteristics at the local level. The local is reliant on the global in that it is the positive or negative result of decisions taken and policies applied at a higher level. But the local level also concentrates all the consequences of improper management at the global level while contributing, through its errors and positive behaviors, to changes at the global level.
ANNs and geographical information for urban analysis evidence from the European FP7 SECOA Project / Montanari, Armando. - In: ARCHEOLOGIA E CALCOLATORI. - ISSN 1120-6861. - STAMPA. - supplemento 6:(2014), pp. 131-151.
ANNs and geographical information for urban analysis evidence from the European FP7 SECOA Project
MONTANARI, ARMANDO
2014
Abstract
This article looks at two elements of geographic analysis applied to urban processes – GISystems, software for analyzing geographic information, and GIScience, theories and methods to interpret geographic information. Self-Organizing Maps (SOM) are a type of Artificial Neural Network (ANN), initially studied by Kohonen, Hynninen, Kangas & Laaksonen (1995), capable of automatically organizing data within an artificial space in such a way as to group together data with similar characteristics while distancing dissimilar data. They are therefore able to preserve the topological relationships of archived data and produce output maps that can be analyzed using a GISystem. Because the “Solutions for environmental contrasts in coastal areas. Global Change, human mobility and sustainable urban development” project (SECOA) used data originating from the experiences of the natural and social sciences, the data collection and processing stages were complex. Complexity is characteristic of our society and, to explain phenomena related to us, it is increasingly necessary to analyze all the available data, whatever their origin: if these phenomena exist, they are relevant in one way or another to the formation of reality. So they must be identified and measured using the expedients that make it possible to subsequently include them in the rest of the information. SECOA considered two levels of phenomena: global ones such as climate change and human mobility, and local ones that emerge in the form of conflicts, the dynamics of urban growth, and natural, environmental, social and economic characteristics at the local level. The local is reliant on the global in that it is the positive or negative result of decisions taken and policies applied at a higher level. But the local level also concentrates all the consequences of improper management at the global level while contributing, through its errors and positive behaviors, to changes at the global level.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
