A Forecast of Geohazard and Factors Influencing Geohazard Using Transfer Learning

Geohazard is an ecological destruction problem that exists in various parts of the universe. Geohazard destroys the complete ecosystem. Geohazard results in both human and economic loss. In India, these Geohazards create an impact of 2% of loss in domestic products and 12% of economic loss. In the advancement of technology at various eras, various methodologies were implemented to predict the Geohazard. The approaches start include (a) Steel sheets technique (b) Installation of sensors (Fiber optic and Electrical Sensor) in expected place (c) Machine learning models (d) Time series analysis (e) Basic neural network structure, etc. The problem faced by conventional approaches are (a) large volume of data, (b) Satellite image-based data (c) Radar covers a small area, etc. Deep learning is the cutting-edge technology that addresses the problems faced by the traditional approaches in effectively. The usage of architectures in deep learning provides the solution for Geohazard. The proposed work implements a novel approach “Transfer learning approaches for effective prediction along with factors influencing Geohazard”. VGG16 is a state-of-art technique for predicting the images more precisely with an accuracy of 80% in recognizing the occurrence of Geohazard. The various factors that influence the Geohazard are identified using Correlation mapping.