Comparison Study of Classification Model for Forest Cover Type Prediction

Public Deposited
Analytics
Request Version for Screen Reader
Last Modified
  • July 28, 2023
Creator
  • Meng, Lingqi
    • Affiliation: School of Information and Library Science
Abstract
  • Forest is one of the most important natural resource that correlate to biodiversity, climate, geochemical aspect. Forest cover type is a dominant kind of tree cover in certain area. This study is addressing a multiclass classification problem of forest cover type via supervised learning method. The cartographic feature of forest will be retrieved from dataset of Roosevelt National forest. There exist seven types of trees and fifty-five environmental features in the area. Six machine learning models, K-NN, linearSVC, logisticRegression, GaussianNB, DecisionTree and RandomForest classifier are implemented and compared. K-NN algorithm gains the best prediction accuracy and best overall performance over all the classes.
Date of publication
DOI
Resource type
Rights statement
  • In Copyright
Advisor
  • Arguello, Jaime
Degree
  • Master of Science in Information Science
Graduation year
  • 2019
Language

Relations

Parents:

This work has no parents.

In Collection:

Items