Impact of Data Analytics in Agriculture: Landscape Approach for Sustainable Land Use

This paper presents an analysis of Bulgaria, which seeks to classify regions into plain, hill, and mountain areas using the Random Forest predictive model. The aim is for the investment in an economic entity to be placed in the right area of ​​relief and to be able to maximize both economic utility and profit. Following the automated classification, by using predictive algorithms, it is wanted to create an analysis of data on the number of firms in the main regions associated with key areas of relief (plain, hill, mountain). The paper aims to create the basis of an automated management system, to be achieved by using predictive techniques, such as machine learning—Random Forest. The research can be used in future analyses, in order to reduce the risk costs that new businesses have when deciding on the location of the economic unit and the most sustainable use of agricultural land.