Flag Counter
AKILLI SİSTEMLER VE UYGULAMALARI DERGİSİ
JOURNAL OF INTELLIGENT SYSTEMS WITH APPLICATIONS
J. Intell. Syst. Appl.
E-ISSN: 2667-6893
Creative Commons License This work is licensed under a Creative Commons Attribution 4.0 International License.

Comparison of Artificial Neural Networks and Genetic Programming Methods For Activity Recognition

Aktivite Tanımada Yapay Sinir Ağları ve Genetik Programlama Yöntemlerinin Karşılaştırılması

How to cite: Erdaş B, Aşuroğlu T, Açıcı K, Oğul H. Comparison of artificial neural networks and genetic programming methods for activity recognition. Akıllı Sistemler ve Uygulamaları Dergisi (Journal of Intelligent Systems with Applications) 2018; 1(1): 62-66. DOI: 10.54856/jiswa.201805019

Full Text: PDF, in Turkish.

Total number of downloads: 585

Title: Comparison of Artificial Neural Networks and Genetic Programming Methods For Activity Recognition

Abstract: With the widespread use of wearable sensors, the processing of raw data obtained from sensors has led to widely-used solutions to the problem of activity recognition. In this context, it is aimed to compare the performance of artificial neural network methods (ANN, RBFNN) and genetic programming (GP) methods over time, frequency and wavelet features extracted from the accelerometer data. The most successful classification performance achieved was 75.09% using 31 neurons in the hidden layer of the multilayer perceptron, using time attributes.

Keywords: Artificial neural networks; genetic programming; activity recognition; sensor


Başlık: Aktivite Tanımada Yapay Sinir Ağları ve Genetik Programlama Yöntemlerinin Karşılaştırılması

Özet: Giyilebilir sensörlerin yaygınlaşmasıyla beraber sensörlerden elde edilen ham verilerin işlenmesiyle aktivite tanıma problemine getirilen çözümler yaygınlaşmaya başlamıştır. Bu bağlamda literatürde çeşitli uygulamalar olmakla beraber, bu çalışmada aktivite tanımada popüler olarak kullanılan yapay sinir ağı yöntemleri (ANN, RBFNN) ve genetik programlama (GP) yönteminin ivmeölçer verisinden çıkarılan zaman, frekans ve dalgacık (wavelet) öznitelikleri üzerinden performans karşılaştırılması hedef alınmıştır. Bahsi geçen çalışma kapsamında alınan sonuçlara bakıldığında, ulaşılan en başarılı sınıflandırma performansı zaman öznitelikleri kullanılarak, çok katmanlı perseptronun ara katmanında 31 nöron kullanılması ile %75.09 olarak elde edilmiştir.

Anahtar kelimeler: Yapay sinir ağları; genetik programlama; aktivite tanıma; sensör


Bibliography:
  • Fang H, He L. BP neural network for human activity recognition in smart home. In International Conference on Computer Science and Service System, August 11-13, 2012, Nanjing, China.
  • Rodrigues LM, Mestria M. Classification methods based on Bayes and neural networks for human activity recognition. In 12th International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery (ICNC-FSKD), August 13-15, 2016, Changsha, China.
  • Mehr HD, Polat H, Cetin A. Resident activity recognition in smart homes by using artificial neural networks. In 4th International Smart Grid Congress and Fair (ICSG), April 20-21, 2016, Istanbul, Turkey.
  • Zeng W, Wang C. Human gait recognition via deterministic learning. Neural Networks 2012; 35: 92-102.
  • Sood M, Bhooshan SV. Automatic processing of EEG signals for seizure detection using soft computing techniques. In International Conference on Recent Advances and Innovations in Engineering, May 09-11, 2014, Jaipur, India.
  • Xie F, Song A, Ciesielski V. Genetic programming based activity recognition on a smartphone sensory data benchmark. In Congress on Evolutionary Computation (CEC), July 6-11, 2014, Beijing, China.