Skip to main content

Advertisement

SpringerLink
Log in

Feature selection and human arm activity classification using a wristband

  • Original Article
  • Published:
Journal of Reliable Intelligent Environments Aims and scope Submit manuscript

Abstract

Activity monitoring is a core application for wristbands and, consequently, all the top selling brands (Xiaomi, Apple, Huawei, Fitbit, and Samsung) incorporate accelerometers as a core movement sensor. Applications range from sports to fitness, supported by algorithms that analyze the sensor data. Thus, there are significant benefits to be accrued from improving the activity classification performance of wrist-worn activity monitors, a goal that this study seeks to address. Further to achieving this goal, this paper presents research which investigates the potential for improving strategies and algorithms used in data pre-processing and model training/testing, for wrist-worn accelerometer sensing. To those ends we investigate different techniques for data sampling frequency, feature ranking, feature scaling and sub-feature sets selection, as well as model selection strategies based on a set of neural network, support vector machine, and Gaussian Naïve Bayes classification algorithms. We explore the effects of different model training and testing strategies, and compare three models trained with different datasets organized by personalization, partial mixing, and full mixing from multiple subjects. Their relative performance is then compared based on different test datasets, which are personalized, mixed with pre-specified training subjects, and non-pre specified (unseen/new) subjects, respectively. Moreover, a novel plurality voting mechanism was explored as a means to adjust the prediction result during the model testing stage. Finally, the paper concludes by presenting the main finding of the research which are that the most robust and reliable performance for human activity classification can be obtained by combining a personalized model with a plurality voting mechanism.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others

References

  1. Chelsey D (2019) IDC: Q2 wrist-worn wearable shipments up 29%. Circuits Assembly Online Magazine. https://circuitsassembly.com/ca/editorial/menu-news/32081-idc-q2-wrist-worn-wearable-shipments-up-29.html. Accessed 26 June 2021

  2. Ramon L (2019) Wrist-worn wearables maintain a strong growth trajectory in Q2 2019, According to IDC. https://www.businesswire.com/news/home/20190912005263/en/Wrist-Worn-Wearables-Maintain-Strong-Growth-Trajectory-Q2. Accessed 26 June 2021

  3. Aroganam G, Manivannan N, Harrison D (2019) Review on wearable technology sensors used in consumer sport applications. Sensors 19(9):1983

    Article  Google Scholar 

  4. Chen C, Kehtarnavaz N, Jafari R (2014) A medication adherence monitoring system for pill bottles based on a wearable inertial sensor. In: Engineering in medicine and biology society (EMBC), 36th annual international conference of the IEEE, pp 4983–4986

  5. Chen C, Liu K, Jafari R, Kehtarnavaz N (2014) Home-based senior fitness test measurement system using collaborative inertial and depth sensors. In: Engineering in medicine and biology society (EMBC), international conference of the IEEE, pp 4135–4138

  6. Cornacchia M, Ozcan K, Zheng Y, Velipasalar S (2017) A survey on activity detection and classification using wearable sensors. IEEE Sens J 17(2):386–403

    Article  Google Scholar 

  7. Mukhopadhyay SC (2015) Wearable sensors for human activity monitoring: a review. IEEE Sens J 15(3):1321–1330

    Article  Google Scholar 

  8. Rostami M, Berahmand K, Nasiri E, Forouzande S (2021) Review of swarm intelligence-based feature selection methods. Eng Appl Artif Intell 100:104210

    Article  Google Scholar 

  9. Tubishat M, Ja’afar S, Alswaitti M, Mirjalili S, Idris N, Ismail MA, Omar MS (2021) Dynamic salp swarm algorithm for feature selection. Expert Syst Appl 164:113873

    Article  Google Scholar 

  10. Quiroz JC, Banerjee A, Dascalu SM, Lau SL (2017) Feature selection for activity recognition from smartphone accelerometer data. Intell Autom Soft Comput 1–9

  11. Xue B, Zhang M, Browne WN, Yao X (2016) A survey on evolutionary computation approaches to feature selection. IEEE Trans Evol Comput 20(4):606–626

    Article  Google Scholar 

  12. Rodgers J, Nicewander WA (1988) Thirteen ways to look at the correlation coefficient. Am Stat 42(1):59–66

    Article  Google Scholar 

  13. Ma L, Li M, Gao Y, Chen T, Ma X, Qu L (2017) A novel wrapper approach for feature selection in object-based image classification using polygon-based cross-validation. IEEE Geosci Remote Sens Lett 14(3):409–413

    Article  Google Scholar 

  14. Yan K, Zhang D (2015) Feature selection and analysis on correlated gas sensor data with recursive feature elimination. Sens Actuators B Chem 212:353–363

    Article  Google Scholar 

  15. Kukreja SL, Löfberg J, Brenner MJ (2006) A least absolute shrinkage and selection operator (LASSO) for nonlinear system identification. IFAC Proc Vol 39(1):814–819

    Article  Google Scholar 

  16. Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, Blondel M, Prettenhofer P, Weiss R, Dubourg V, Vanderplas J (2011) Scikit-learn: machine learning in Python. J Mach Learn Res 12(Oct):2825–2830

    MathSciNet  MATH  Google Scholar 

  17. Grus J (2015) Data science from scratch: first principles with Python. O’Reilly Media, Inc.

  18. Ioffe S, Szegedy C (2015) Batch normalization: accelerating deep network training by reducing internal covariate shift. In: International conference on machine learning, pp 448–456

  19. Busso C, Mariooryad S, Metallinou A, Narayanan S (2013) Iterative feature normalization scheme for automatic emotion detection from speech. IEEE Trans Affect Comput 4(4):386–397

    Article  Google Scholar 

  20. Mohamad IB, Usman D (2013) Standardization and its effects on K-means clustering algorithm. Res J Appl Sci Eng Technol 6(17):3299–3303

    Article  Google Scholar 

  21. Gao L, Bourke AK, Nelson J (2014) Evaluation of accelerometer based multi-sensor versus single-sensor activity recognition systems. Med Eng Phys 36(6):779–785

    Article  Google Scholar 

  22. Tchuente F, Baddour N, Lemaire ED (2020) Classification of aggressive movements using smartwatches. Sensors 20(21):6377

    Article  Google Scholar 

  23. Eibe F, Hall MA, Ian H (2016) The WEKA Workbench. Online appendix for "data mining: practical machine learning tools and techniques", 4th edn. Morgan Kaufmann

  24. Kotsiantis SB (2007) Supervised machine learning: a review of classification techniques. Informatica 31:249–268

    MathSciNet  MATH  Google Scholar 

  25. Yang J (2009) Toward physical activity diary: motion recognition using simple acceleration features with mobile phones. In: Proceedings of the 1st international workshop on Interactive multimedia for consumer electronics, pp 1–10

  26. Namsrai E, Munkhdalai T, Li M, Shin J, Namsrai O, Ryu KH (2013) A feature selection-based ensemble method for arrhythmia classification. J Inf Process Syst 9(1):31–40

    Article  Google Scholar 

  27. Ngo TT, Makihara Y, Nagahara H et al (2015) Similar gait action recognition using an inertial sensor. Pattern Recogn 48(4):1289–1301

    Article  Google Scholar 

  28. Jadhav SD, Channe HP (2016) Comparative study of K-NN, naive Bayes and decision tree classification techniques. Int J Sci Res (IJSR) 5(1):1842–1845

    Article  Google Scholar 

  29. Weiss GM, Lockhart J (2012) The impact of personalization on smartphone-based activity recognition. In: Workshops at the twenty-sixth AAAI conference on artificial intelligence

  30. Tapia EM, Intille SS, Haskell W, Larson K, Wright J, King A, Friedman R (2007) Real-time recognition of physical activities and their intensities using wireless accelerometers and a heart rate monitor. In: 2007 11th IEEE international symposium on wearable computers, pp 37–40

  31. Ferrari A, Micucci D, Mobilio M, Napoletano P (2021) Trends in human activity recognition using smartphones. J Reliab Intell Environ 7(3):189–213

    Article  Google Scholar 

  32. Burns DM, Whyne CM (2020) Personalized activity recognition with deep triplet embeddings. http://arxiv.org/abs/2001.05517

  33. Axiviy X3 accelerometers website, https://axivity.com/. Accessed 20 June 2021

  34. Murphy C (2017) Choosing the most suitable MEMs accelerometer for your application—part 2. Analog Dialogue 51(11):1–6

    Google Scholar 

  35. Liu H, Motoda H (1999) Feature extraction construction and selection: a data mining perspective. J Am Stat Assoc 94(448):014004

    Article  Google Scholar 

  36. Ferrari A, Micucci D, Marco M, Napoletano P (2019) Hand crafted features vs residual networks for human activities recognition using accelerometer. In: Proceedings of the IEEE international symposium on consumer technologies (ISCT), pp 153–156.

  37. Quigley B, Donnelly M, Moore G, Galway L (2018) A comparative analysis of windowing approaches in dense sensing environments. In: Multidisciplinary digital publishing institute proceedings, vol 2, no 19, p 1245

  38. Banos O, Galvez JM, Damas M, Pomares H, Rojas I (2014) Window size impact in human activity recognition. Sensors 14(4):6474–6499

    Article  Google Scholar 

Download references

Funding

This work was supported by Hebei Science and Technology Department, Innovation Capability Improvement Plan Project, Grant Numbers 21557611K and Hebei Scientific Research Platform Construction project, Grant Nos. SG20182058 and SZX2020033.

Author information

Authors and Affiliations

  1. Shijiazhuang University, Shijiazhuang, China

    Shumei Zhang, Xiangping An, Liying Duan & Hua Liu

  2. Hebei Key Laboratory of IoT Blockchain Integration, Shijiazhuang, China

    Shumei Zhang

  3. School of Computer Science and Electronic Engineering, Essex University, Essex, UK

    Vic Callaghan

Authors
  1. Shumei Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Vic Callaghan
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Xiangping An
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Liying Duan
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Hua Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Shumei Zhang.

Ethics declarations

Conflict of interest

The authors have no competing interests to declare that are relevant to the content of this article.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zhang, S., Callaghan, V., An, X. et al. Feature selection and human arm activity classification using a wristband. J Reliable Intell Environ 8, 285–298 (2022). https://doi.org/10.1007/s40860-022-00181-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s40860-022-00181-6

Keywords

Search

Navigation