DocumentCode
629539
Title
Performance evaluation of feature selection algorithms on human activity classification
Author
Tulum, Gokalp ; Artug, N. Tugrul ; Bolat, B.
Author_Institution
Electr. & Electron. Eng., Yeni Yuzyil Univ., Istanbul, Turkey
fYear
2013
fDate
19-21 June 2013
Firstpage
1
Lastpage
4
Abstract
In this work, four human activities were classified by using multi layer perceptron and k-nearest neighbours algorithm. Due to mass amount of data, two different feature selection methods, which are ReliefF and t-score, were applied to the data. The best result is obtained as 97.6% with 51 features selected by ReliefF.
Keywords
feature extraction; image classification; multilayer perceptrons; object recognition; ReliefF; feature selection algorithms; human activity classification; human activity recognition; k-nearest neighbours; multilayer perceptron algorithm; performance evaluation; t-score; Accuracy; Classification algorithms; Filtering algorithms; Sensor phenomena and characterization; Tactile sensors; Training; Feature selection; ReliefF; human activity detection; t-score;
fLanguage
English
Publisher
ieee
Conference_Titel
Innovations in Intelligent Systems and Applications (INISTA), 2013 IEEE International Symposium on
Conference_Location
Albena
Print_ISBN
978-1-4799-0659-8
Type
conf
DOI
10.1109/INISTA.2013.6577634
Filename
6577634
Link To Document