Mahalanobis uzaklığı tabanlı aykırı değer bulma ve ReliefF öznitelik seçimine dayalı bir makine öğrenmesi yaklaşımı ile akıllı telefon verileri üzerinden stres tespiti

Öz

Stres kişinin odaklanması, uyanık kalması ve tetikte olması durumlarında fayda sağlamaktadır. Fakat yüksek dozda strese maruz kalmak kişinin sağlığına zarar vermektedir. Bu nedenle stresin tespit edilip en kısa sürede rahatlamaya geçilmesi önemlidir. Bu çalışmada, akıllı telefondan elde edilen dokunmatik panel, yerçekimi, doğrusal ivme ve jiroskop verileri ile yazma davranışları incelenmiştir. Elde edilen sonuçlardan yazma davranışları ile kişilerin stres seviyeleri arasında bir bağlantı olduğu görülmüştür. Bu kapsamda genişletilmiş bir veri kümesi oluşturulmuştur. Stresin daha yüksek doğrulukta tespit edilebilmesi için Mahalanobis uzaklığı tabanlı bir aykırı veri tespiti yaklaşımı uygulanmıştır. Devamında, verimli özniteliklerin tespit edilerek sınıflandırma gerçekleştirilmesi için ReliefF öznitelik seçimi yöntemi ve makine öğrenmesi teknikleri kombine edilerek bir yapı oluşturulmuştur. Aykırı verilerin temizlenerek elde edilen sonuçlar, oluşturulan yapıların yüksek doğrulukta başarı yakaladığını göstermiştir. Ek olarak aykırı veri tespiti ve temizliği, sınıflandırma başarısını 1.77 puan artırmıştır.

Anahtar Kelimeler

• Stres tespiti
• Mahalanobis uzaklığı
• Öznitelik değerlendirme
• Akıllı telefon verileri
• Yazma davranışı

Stress detection on smartphone data with a machine learning approach based on Mahalanobis distance-based outlier finding and ReliefF feature selection

Abstract

Stress is beneficial when a person is focused, awake and alert. However, exposure to high doses of stress harms a person's health. For this reason, it is important to detect stress and begin relief as soon as possible. In this study, soft keyboard typing behaviors with touchscreen panel, gravity, linear acceleration, and gyroscope data obtained from smartphones were examined. It was observed that there was a correlation between the results obtained and typing behaviors and the stress levels of individuals. In this context, an expanded data set was created. In order to detect stress with higher accuracy, a Mahalanobis distance-based outlier detection approach was applied. Subsequently, a structure was created by combining the ReliefF feature selection method and machine learning techniques to identify efficient features and perform classification. The results obtained by cleaning outlier data showed that the created structures achieved success with high accuracy. In addition, outlier detection and cleaning increased the classification success by 1.77 points.

Keywords

• Stress detection
• Mahalanobis distance
• Feature evaluation
• Smartphone data
• Typing behavior

Kaynakça

1. [1] Sano A, Picard RW. “Stress recognition using wearable sensors and mobile phones”. Humaine Association Conference on Affective Computing and Intelligent Interaction, Geneva, Switzerland, 2-5 September 2013.
2. [2] Can YS, Arnrich B, Ersoy C. “Stress detection in daily life scenarios using smart phones and wearable sensors: A survey”. Journal of Biomedical Informatics, 92, 1-22, 2019.
3. [3] Balli S, Sağbaş EA, Peker M. “Human activity recognition from smart watch sensor data using a hybrid of principal component analysis and random forest algorithm”. Measurement and Control, 52(1-2), 37-45, 2019.
4. [4] Güven Z, Diri B, Çakaloğlu T. “Comparison of n-stage latent dirichlet allocation versus other topic modeling methods for emotion analysis”. Journal of the Faculty of Engineering and Architecture of Gazi University, 35(4), 2135-2145, 2020.
5. [5] Gokalp O, Tasci E, Ugur A. “A novel wrapper feature selection algorithm based on iterated greedy metaheuristic for sentiment classification”. Expert Systems with Applications, 146, 1-10, 2020.
6. [6] Sağbaş EA, Korukoglu S, Balli S. “Stress detection via keyboard typing behaviors by using smartphone sensors and machine learning techniques”. Journal of Medical Systems, 44(4), 1-12, 2020.
7. [7] Lu H, Frauendorfer D, Rabbi M, Mast MS, Chittaranjan GT, Campbell AT, Gatica-perez D, Choudhury T. “Stresssense: detecting stress in unconstrained acoustic environments using smartphones”. ACM Conference on Ubiquitous Computing, Pittsburg, USA, 5-8 September 2012.
8. [8] Wang R, Chen F, Chen Z, Li T, Harari G, Tignor S, Zhou X, Ben-Zeev D, Campbell AT. “StudentLife: assessing mental health, academic performance and behavioral trends of college students using smartphones”. ACM International Joint Conference on Pervasive and Ubiquitous Computing, Seattle Washington, 3-14, September 2014.

9. [9] Garcia-Ceja E, Osmani V, Mayora O. “Automatic stress detection in working environments from smartphones’ accelerometer data: a first step”. IEEE Journal of Biomedical and Health Informatics, 20(4), 1053-1060, 2015.
10. [10] Vildjiounaite E, Kallio J, Kyllönen V, Nieminen M, Määttänen I, Lindholm M, Mäntyjärvi J, Gimel’farb G. “Unobtrusive stress detection on the basis of smartphone usage data”. Personal and Ubiquitous Computing, 22(4), 671-688, 2018.
11. [11] Ferdous R, Osmani V, Mayora O. “Smartphone app usage as a predictor of perceived stress levels at workplace”. International Conference on Pervasive Computing Technologies for Healthcare, Istanbul, Turkey, 20-23 May 2015.
12. [12] Stütz T, Kowar T, Kager M, Tiefengrabner M, Stuppner M, Blechert J, Wilhelm FH, Ginzinger S. “Smartphone based stress prediction”. UMAP: International Conference on User Modeling, Adaptation and Personalization, Dublin, Ireland, 29 June - 3 July 2015.
13. [13] Muaremi A, Arnrich B, Tröster G. “Towards measuring stress with smartphones and wearable devices during workday and sleep”. BioNanoScience, 3(2), 172-183, 2013.
14. [14] Bauer G, Lukowicz P. “Can smartphones detect stressrelated changes in the behaviour of individuals?”. International Conference on Pervasive Computing and Communications Workshops, Lugano, Switzerland, 19-23 March 2012.
15. [15] Sysoev M, Kos A, Pogačnik M. “Noninvasive stress recognition considering the current activity”. Personal and Ubiquitous Computing, 19(7), 1045-1052, 2015.
16. [16] Gjoreski M, Gjoreski H, Lutrek M, Gams M. “Automatic detection of perceived stress in campus students using smartphones”. International Conference on Intelligent Environments, Prague, Czech Republic, 15-17 July 2015.
17. [17] Bogomolov A, Lepri B, Ferron M, Pianesi F, Pentland A. “Daily stress recognition from mobile phone data, weather conditions and individual traits”. ACM international conference on Multimedia, Orlando, Florida, USA, 3-7 November 2014.
18. [18] Ciman M, Wac K, Gaggi O. “iSenseStress: assessing stress through human-smartphone interaction analysis”. International Conference on Pervasive Computing Technologies for Healthcare, Istanbul, Turkey, 20-23 May 2015.
19. [19] Ghosh S, Ganguly N, Mitra B, De P. “Tapsense: Combining self-report patterns and typing characteristics for smartphone based emotion detection”. International Conference on Human-Computer Interaction with Mobile Devices and Services, Vienna, Austria, 4-7 September 2017.
20. [20] Ghosh S, Sahu S, Ganguly N, Mitra B, De P. “EmoKey: an emotion-aware smartphone keyboard for mental health monitoring”. International Conference on Communication Systems & Networks, Bengaluru, India, 7-11 January 2019.
21. [21] Gao Y, Bianchi-Berthouze N, Meng H.“What does touch tell us about emotions in touchscreen-based gameplay?”. ACM Transactions on Computer-Human Interaction, 19(4), 1-30, 2012.
22. [22] Kim HJ, Choi YS. “Exploring emotional preference for smartphone applications”. IEEE Consumer Communications and Networking Conference, Las Vegas, NV, USA, 14-17 January 2012.
23. [23] Ciman M, Wac K. “Individuals’ stress assessment using human-smartphone interaction analysis”. IEEE Transactions on Affective Computing, 9(1), 51-65, 2016.
24. [24] Lee H, Choi YS, Lee S, Park IP. “Towards unobtrusive emotion recognition for affective social communication”. IEEE Consumer Communications and Networking Conference, Las Vegas, NV, USA, 14-17 January 2012.
25. [25] Sağbaş EA, Ballı S. “Akıllı telefon sensörlerinin kullanimi ve ham sensör verilerine erişim”. XVII. Akademik Bilişim Konferansı, Eskişehir, Türkiye, 4-6 Şubat 2015.
26. [26] Stroop JR. “Studies of interference in serial verbal reactions”. Journal of Experimental Psychology, 18(6), 643-662, 1935.
27. [27] Sağbaş E.A, Ballı S. “Akıllı telefon algılayıcıları ve makine öğrenmesi kullanılarak ulaşım türü tespiti”. Pamukkale Üniversitesi Mühendislik Bilimleri Dergisi, 22(5), 376-383, 2016.
28. [28] Sağbaş, EA, Ballı, S. “Akıllı saat algılayıcıları ile insan hareketlerinin sınıflandırılması”. Süleyman Demirel Üniversitesi Fen Bilimleri Enstitüsü Dergisi, 21(3), 980-990, 2017.
29. [29] Witten IH, Frank E. “Data mining: practical machine learning tools and techniques with Java implementations”. ACM Sigmod Record, 31(1), 76-77, 2002.
30. [30] Ghorbani H. “Mahalanobis distance and its application for detecting multivariate outliers”. Facta Universitatis, Series: Mathematics and Informatics, 34(3), 583-95, 2019.
31. [31] Yuksel AS, Senel, FA, Cankaya IA. “Classification of soft keyboard typing behaviors using Mobile device sensors with machine learning”. Arabian Journal for Science and Engineering, 44(4), 3929–3942, 2019.
32. [32] Peker M, Arslan A, Şen B, Çelebi FV, But A. “A novel hybrid method for determining the depth of anesthesia level: Combining ReliefF feature selection and random forest algorithm (ReliefF+ RF)”. International Symposium on Innovations in Intelligent SysTems and Applications, Madrid, Spain, 2-4 September 2015.
33. [33] Ballı S, Sağbaş EA. “Diagnosis of transportation modes on mobile phone using logistic regression classification”. IET Software, 12(2), 142-151, 2018.
34. [34] Kononenko I. “Estimating attributes: analysis and extensions of Relief”. ECML: European Conference on Machine Learning, Catania, Italy, 6–8 April 1994.
35. [35] Robnik-Šikonja M, Kononenko I. “Theoretical and empirical analysis of ReliefF and RReliefF”. Machine Learning, 53(1), 23-69, 2003.