Evaluating existing manually constructed natural landscape classification with a machine learning-based approach

Some landscape classifications officially determine financial obligations; thus, they must be objective and precise. We presume it is possible to quantitatively evaluate existing manually constructed classifications and correct them if necessary. One option for achieving this goal is a machine learn...

Full description

Bibliographic Details
Main Authors: Rok Ciglič, Erik Štrumbelj, Rok Češnovar, Mauro Hrvatin, Drago Perko
Format: Article
Language:English
Published: University of Maine 2019-06-01
Series:Journal of Spatial Information Science
Subjects:
Online Access:http://josis.org/index.php/josis/article/view/464

Similar Items