Geospatial Data Analytics- A Deep Learning Perspective

Abstract

To visualize high-dimensional geospatial data has achieved much importance in last decades. But to analyze it, the technologies used of machine learning are not so convincing and thus it is high time to switch to a sub-domain of machine learning called deep learning, which has gained popularity because of its accuracy in processing and analyzing high-dimensional data. The convergence of deep learning with geospatial data analytics shall prove to be a boon to those who actually has a need to predict specific outputs over geospatial data. In this paper, we have presented some geospatial data generated using the GIS (Geographic Information Systems) technology and proposed ways to implement deep learning over these data. GIS technology is a mapping technology which signifies high-dimensional geospatial data and our aim is to propose a model where GIS converges with highly efficient deep neural networks such as CNN (Convolutional Neural Network), RNN (Recurrent Neural Network) and LSTM (Long Short-Term Memory Network). We have provided geospatial as well as statistical results in this paper to visualize practically the GIS technology. This paper further provides future scope of the proposed model which shall present the challenges that needs to be tackled in future and its applicability in many relevant domains.