A Comparison of Decision Tree Algorithms for Indoor User Localization Using Wireless Signal Strength
Localizing users and devices indoors has a wide range of applications. Smart home systems can be used to locate criminals in restricted areas and determine the number of users at an access point. The aim of this study is to determine the location of users indoors using wireless signal strength as we...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Istanbul University Press
2022-12-01
|
| Series: | Acta Infologica |
| Subjects: | |
| Online Access: | https://cdn.istanbul.edu.tr/file/JTA6CLJ8T5/DB504088DE3447F0A4F6CF4C19760BB4 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Localizing users and devices indoors has a wide range of applications. Smart home systems can be used to locate criminals in restricted areas and determine the number of users at an access point. The aim of this study is to determine the location of users indoors using wireless signal strength as well as the best decision tree classification algorithm that can be used in monitoring devices that will be designed. For this purpose, the study uses 12 different algorithms and compares their performances by conducting a performance analysis. The study uses 10- fold cross validation as the performance analysis method. While evaluating the performance, the algorithms’ classification performance were compared before and after the cross-validation. Due to the study using a balanced dataset, the performance metrics used for classifying balanced datasets have bene preferred in the performance analysis. As a result of the analysis, the random forest algorithm was observed to have achieved the best performance. All metric values calculated before and after the cross-validation of the random forest algorithm were higher than those for the other algorithms. |
|---|---|
| ISSN: | 2602-3563 |