The article proposes a solution for organizing a storage of artificial neural networks in a digital spatial data infrastructure system. Based on the analysis of world experience, a register of key storage cases was created, which made it possible to create an effective solution for analyzing large arrays of spatial data. The structure of the neural network sets the format of the input data and the type of the output signal. It is shown that the use of neural networks for solving design problems requires dividing the storage of the ontological model into machine learning, data and task modules. The introduction of deep learning models into the repository will allow not only to form an ANN system capable of solving urgent problems in the field of analysis of different types of big data, but also to solve the problem of choosing an effective model by building a system of recommendations that optimize the choice of algorithms.

No abstract

Objectives. The paper aimed to develop and validate a neural network model for spatial data analysis. The advantage of the proposed model is the presence of a large number of degrees of freedom allowing its flexible configuration depending on the specific problem. This development is part of the knowledge base of a deep machine learning model repository including a dynamic visualization subsystem based on adaptive web interfaces allowing interactive direct editing of the architecture and topology of neural network models.Methods. The presented solution to the problem of improving the accuracy of spatial data analysis and classification is based on a geosystem approach for analyzing the genetic homogeneity of territorial-adjacent entities of different scales and hierarchies. The publicly available EuroSAT dataset used for initial validation of the proposed methodology is based on Sentinel-2 satellite imagery for training and testing machine learning models aimed at classifying land use/land cover systems. The ontological model of the repository including the developed model is decomposed into domains of deep machine learning models, project tasks and data, thus providing a comprehensive definition of the formalizing area of knowledge. Each stored neural network model is mapped to a set of specific tasks and datasets. Results. Model validation for the EuroSAT dataset algorithmically extended in terms of the geosystem approach allows classification accuracy to be improved under training data shortage within 9% while maintaining the accuracy of ResNet50 and GoogleNet deep learning models.Conclusions. The implemention of the developed model into the repository enhances the knowledge base of models for spatial data analysis as well as allowing the selection of efficient models for solving problems in the digital economy.

The work describes the key principles of the process of building digital spatial data infrastructures for effective decision-making in the management of natural systems and for the sustainable development of the regional economy. The following reference points are considered in detail: increasing the accuracy of the deep learning and neural networks algorithmic and software for the process of analyzing spatial data, developing storage systems for large spatio-temporal data by developing new physical and logical storage models, introducing effective geoportal technologies and developing new architectural patterns for presentation and further dissemination of spatio-temporal using modern web technologies. The plan for working out a scientific problem of development of methods and architectural patterns of storage, analysis and distribution of spatio-temporal data determined the structure of the article. The first section concretizes the criteria of efficiency of information processes in the digital spatial data infrastructure (SDI), the second section discusses algorithmic support of the process of analysis of spatial data, the thirdintegration of spatial data, and finally, the final sectionimplementation and projectoriented use of geoportal systems.