Advanced Methods for Geographers

Advanced Methods for Geographers

Study Cycle: 2




ECTS credit:6

Lecturer(s): asist. Bobovnik Nejc, doc. dr. Rogelj Boštjan, izr. prof. dr. Krevs Marko


- upgrading the knowledge and skills of multivariate statistical methods in geography: principal components and factor analysis, discriminant analysis, their exploratory and predictive use,
- the purpose and characteristics of machine learning / data mining in geography, the relationship between statistics and machine learning, an overview of machine learning algorithms and tools for their implementation,
- basics of the decision tree method: regression and classification trees, selected partitioning methods, binary and multidirectional node partitioning, substitution of predictor variables, measures of predictive accuracy, different approaches to evaluating model predictions, possible approaches to improving the results of the decision tree method,
- use of Python scripts in performing GIS-based tasks: online resources of Python scripts, guided (group) and independent learning and training for using scripts in performing geoinformation tasks,
- written project work (with independently performed examples on the use of multivariate statistical methods, decision trees and own Python script), documentation and research database, presentation in the seminar.