Digital repository of Slovenian research organisations

Show document
A+ | A- | Help | SLO | ENG

Title:Uporaba metod strojnega učenja za preučevanje odnosov med značilnostmi branik in okoljem
Authors:ID Jevšenak, Jernej (Author)
ID Džeroski, Sašo (Author)
ID Levanič, Tom (Author)
Files:.pdf PDF - Presentation file, download (1,18 MB)
MD5: E97089E7EF1BB6098079AF533D74D48C
PID: 20.500.12556/dirros/08b5e5d9-4507-43f2-947e-1d470da2a673
 
URL URL - Source URL, visit https://doi.org/10.20315/ASetL.114.2
 
Language:Slovenian
Typology:1.01 - Original Scientific Article
Organization:Logo SciVie - Slovenian Forestry Institute
Abstract:Različne študije so pokazale, da lahko z nelinearnimi metodami bolje opišemo (modeliramo) odnos med branikami in okoljem. V naši študiji smo primerjali (multiplo) linearno regresijo (MLR) in štiri nelinearne metode strojnega učenja: modelna drevesa (MT), ansambel bagging modelnih dreves (BMT), umetne nevronske mreže (ANN) in metodo naključnih gozdov (RF). Za primerjavo teh metod modeliranja smo uporabili štiri množice podatkov. Natančnost naučenih modelov smo ocenili z metodo 10-kratnega prečnega preverjanja (ang. 10-fold cross-validation) na naši množici in preverjanjem na dodatni testni množici. Na vseh množicah smo dobili boljše statistične kazalce za nelinearne metode s področja strojnega učenja, s katerimi lahko pojasnimo večji delež variance oz. dobimo manjšo napako. Nobena metoda se ni pokazala kot najboljša v vseh primerih, zato je smiselno predhodno primerjati več različnih metod in nato uporabiti najprimernejšo, npr. za rekonstrukcijo klime.
Keywords:strojno učenje, primerjava metod, dendroklimatologija, umetne nevronske mreže, modelna drevesa, ansambel modelnih dreves, naključni gozdovi, linearna regresija
Publication status:Published
Publication version:Version of Record
Year of publishing:2017
Number of pages:str. 21-29
Numbering:Vol. 114
PID:20.500.12556/DiRROS-8175 New window
UDC:630*52:630*11(045)=163.6
ISSN on article:2335-3112
DOI:10.20315/ASetL.114.2 New window
COBISS.SI-ID:4998310 New window
Note:Besedilo v slov.;
Publication date in DiRROS:21.02.2018
Views:5298
Downloads:3290
Metadata:XML RDF-CHPDL DC-XML DC-RDF
:
Copy citation
  
Share:Bookmark and Share


Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Record is a part of a journal

Title:Acta Silvae et Ligni
Publisher:Gozdarski inštitut Slovenije, založba Silva Slovenica, Biotehniška fakulteta, Oddelek za gozdarstvo in obnovljive gozdne vire, Biotehniška fakulteta, Oddelek za lesarstvo
ISSN:2335-3112
COBISS.SI-ID:266761216 New window

Licences

License:CC BY-NC-ND 2.5 SI, Creative Commons Attribution-NonCommercial-NoDerivs 2.5 Slovenia
Link:https://creativecommons.org/licenses/by-nc-nd/2.5/si/deed.en
Description:You are free to reproduce and redistribute the material in any medium or format. You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. You may not use the material for commercial purposes. If you remix, transform, or build upon the material, you may not distribute the modified material. You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.
Licensing start date:21.02.2018

Secondary language

Language:English
Title:On the use of machine learning methods to study the relationships between tree-ring characteristics and the environment
Abstract:Many studies have shown that by using nonlinear methods, the relationship between tree-ring parameters and the environment can be described (modelled) better and in more detail. In our study, (multiple) linear regression (MLR) with four nonlinear machine learning methods are compared: artificial neural networks (ANN), model trees (MT), bagging of model trees (BMT) and random forests of regression trees (RF). To compare the different regression methods, four datasets were used. The performance of the learned models was estimated by using 10-fold cross-validation and an additional hold-out test. For all datasets, better results were obtained by the nonlinear machine learning regression methods, which can explain more variance and yield lower error. However, none of the considered methods outperformed all other methods for all datasets. Therefore, we suggest testing several different methods before selecting the best one, e.g. for climate reconstruction.
Keywords:machine learning, method comparison, dendroclimatology, artificial neural networks, model trees, ensembles of model trees, random forest, linear regression


Collection

This document is a part of these collections:
  1. Acta Silvae et Ligni

Back