Optimal model selection for posture recognition in home-based healthcare

  • Shumei Zhang
  • Paul McCullaghEmail author
  • Chris Nugent
  • Huiru Zheng
  • Matthias Baumgarten
Original Article


This paper investigates optimal model selection for posture recognition. Accuracy and computational time are related to the trained model in a supervised classification. An optimal model selection is important for a reliable activity monitoring system. Conventional guidance on model training uses large instances of randomly selected data in order to characterize the classes. A new approach to the training of a multiclass support vector machine (SVM) model suited to limited training sets such as used in posture recognition is provided. This approach picks a small training set from misclassified data to improve an initial model in an iterative and incremental fashion. In addition, a two step grid-search algorithm is used for the parameters setting. The best parameters were chosen according to the testing accuracy rather than conventional validating accuracy. This new approach for model selection was evaluated against conventional approaches in an activity classification study. Nine everyday postures were classified from a belt-worn smart phone’s accelerometer data. The classification derived from the small training set and the conventional randomly selected training set differed in two aspects: classification performance to new data (85.1% Pick-out small training set vs. 70.3% conventional large training set) and computational efficiency (improved 28%).


Optimal model Posture recognition Accelerometer Multi-class SVM 



The authors acknowledge the support of University of Ulster Vice Chancellor Scholarship Programme, and thank all members of the Smart Environments Research Group for their help with collecting the experimental data. Special thanks are due to Mr W Burns for assisting with the code for data collection using the HTC phone.


  1. 1.
    Abe S (2005) Support vector machines for pattern classification. Springer-Verlag, New YorkGoogle Scholar
  2. 2.
    Abe S, Inoue T (2001) Fast training of support vector machines by extracting boundary data. Artif Neural Netw ICANN 2130:308–313Google Scholar
  3. 3.
    Allen FR, Ambikairajah E, Lovell NH, Celler BG (2006) Classification of a known sequence of motions and postures from accelerometry data using adapted Gaussian mixture models. Physiol Meas 27(10):935–952CrossRefGoogle Scholar
  4. 4.
    Bao L, Intille SS (2004) Activity recognition from user-annotated acceleration data. Pervasive Comput 3001:1–17Google Scholar
  5. 5.
    Boulay B, Brémond F, Thonnat M (2006) Applying 3D human model in a posture recognition system. Pattern Recogn Lett 27(15):1788–1796CrossRefGoogle Scholar
  6. 6.
    Brown EN, Kass RE, Mitra PP (2004) Multiple neural spike train data analysis: state-of-the-art and future challenges. Nat Neurosci 7:456–461CrossRefGoogle Scholar
  7. 7.
    Bussmann J, Martens W, Tulen J, Schasfoort F, VAN DEN BE (2001) Measuring daily behavior using ambulatory accelerometry: the activity monitor. Behav Res Methods Instrum Comput 33(3):349CrossRefGoogle Scholar
  8. 8.
    Bremner D, Demaine E, Erickson J, Iacono J, Langerman S, Morin P, Toussaint G (2005) Output-sensitive algorithms for computing nearest-neighbor decision boundaries. Discrete Comput Geom 33(4):593–604MathSciNetzbMATHCrossRefGoogle Scholar
  9. 9.
    Chang CC, Lin CJ (2001) LIBSVM: a library for support vector machines, 2001) Software available at
  10. 10.
    Cucchiara R, Grana C, Prati A, Vezzani R (2005) Probabilistic posture classification for human-behavior analysis. IEEE Trans Syst Man Cybern Part A 35(1):42–54CrossRefGoogle Scholar
  11. 11.
    Domingos P, Pazzani M (1997) On the the optimality of the simple Bayesian classifier under zero-one loss. Mach Learn 29(2–3):103–130zbMATHCrossRefGoogle Scholar
  12. 12.
    Fletcher R (1981) Practical methods of optimization: constrained optimization. vol 2. John Wiley & Sons, Inc., Somerset, NJ, p 224Google Scholar
  13. 13.
    Foerster F, Smeja M, Fahrenberg J (1999) Detection of posture and motion by accelerometry: a validation study in ambulatory monitoring. Comput Hum Behav 15(5):571–583CrossRefGoogle Scholar
  14. 14.
    Hall M, Frank E (2008) Combining naive Bayes and decision tables, In Proc 21st Florida Artificial Intelligence Research Society Conference, Miami, Florida. AAAI PressGoogle Scholar
  15. 15.
    Harms H, Amft O, Tröster G, Roggen D (2008) Smash: a distributed sensing and processing garment for the classification of upper body postures. Proceedings of the ICST 3rd international conference on body area networks ICST (Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering), pp 22Google Scholar
  16. 16.
    Hsu CW, Lin CJ (2002) A comparison of methods for multiclass support vector machines. IEEE Trans Neural Netw 13(2):415–425CrossRefGoogle Scholar
  17. 17.
    Juang CF, Chang CM (2007) Human body posture classification by a neural fuzzy network and home care system application. IEEE Trans Syst Man Cybern Part A 37(6):984–994MathSciNetCrossRefGoogle Scholar
  18. 18.
    Karhu O, Kansi P, Kuorinka I (1977) Correcting working postures in industry: a practical method for analysis. Appl Ergon 8(4):199–201CrossRefGoogle Scholar
  19. 19.
    Knerr S, Personnaz L, Dreyfus G, Fogelman J, Agresti A, Ajiz M, Jennings A, Alizadeh F, Alizadeh F, Haeberly J (1990) Single-layer learning revisited: a stepwise procedure for building and training a neural network. Optim Methods Softw 1:23–34Google Scholar
  20. 20.
    Koggalage R, Halgamuge S (2004) Reducing the number of training samples for fast support vector machine classification. Neural Inf Process Lett Rev 2(3):57–65Google Scholar
  21. 21.
    Lee S, Park H, Hong S, Lee K, Kim Y (2003) A study on the activity classification using a triaxial accelerometer. Engineering in Medicine and Biology Society, 2003. Proceedings of the 25th Annual International Conference of the IEEEGoogle Scholar
  22. 22.
    Martiskainen P, Järvinen M, Skön JP, Tiirikainen J, Kolehmainen M, Mononen J (2009) Cow behaviour pattern recognition using a three-dimensional accelerometer and support vector machines. Appl Anim Behav Sci 119(1–2):32–38CrossRefGoogle Scholar
  23. 23.
    Mathie MJ, Celler BG, Lovell NH, Coster ACF (2004) Classification of basic daily movements using a triaxial accelerometer. Med Biol Eng Comput 42(5):679–687CrossRefGoogle Scholar
  24. 24.
    Mattmann C, Clemens F, Tröster G (2008) Sensor for measuring strain in textile. Sensors 8(6):3719CrossRefGoogle Scholar
  25. 25.
    Mierswa I, Wurst M, Klinkenberg R, Scholz M, Euler T (2006) YALE: rapid prototyping for complex data mining tasks. Proceedings of the 12th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp 935–940Google Scholar
  26. 26.
    Panini L, Cucchiara R (2003) A machine learning approach for human posture detection in domotics applications. Image Analysis and Processing, 2003. Proceedings. 12th International Conference on, pp 103Google Scholar
  27. 27.
    Ravi N, Dandekar N, Mysore P, Littman ML (2005) Activity recognition from accelerometer data. Proceedings of the National Conference on Artificial Intelligence. Menlo Park, CA; Cambridge, MA; London; AAAI Press; MIT Press; 1999. pp 1541Google Scholar
  28. 28.
    Sohn S, Dagli CH (2002) Advantages of using fuzzy class memberships in self-organizing map and support vector machines. Neural Networks, 2001. Proceedings. IJCNN’01. International Joint Conference on IEEE, pp 1886Google Scholar
  29. 29.
    Sung M, Marci C, Pentland A (2005) Wearable feedback systems for rehabilitation. J Neuroeng Rehabil 2:17CrossRefGoogle Scholar
  30. 30.
    Vapnik V, Lerner A (1963) Pattern recognition using generalized portrait method. Autom Remote Control 24(6):774–780Google Scholar
  31. 31.
    Witten IH, Frank E (2005) Data mining: practical machine learning tools and techniques, 2nd edn. The Morgan Kaufmann series in data management system, ElsevierGoogle Scholar
  32. 32.
    Yang T (2006) Computational verb decision trees. Int J Comput Cogn 4(4):34–46Google Scholar
  33. 33.
    Zhang S, McCullagh P, Nugent C, Zheng H (2009) A theoretic algorithm for fall and motionless detection. Pervasive computing technologies for healthcare, Proceedings of the 3rd Annual IEEE International Conference, pp 1–6Google Scholar

Copyright information

© Springer-Verlag 2010

Authors and Affiliations

  • Shumei Zhang
    • 1
  • Paul McCullagh
    • 1
    Email author
  • Chris Nugent
    • 1
  • Huiru Zheng
    • 1
  • Matthias Baumgarten
    • 1
  1. 1.Computer Science Research Institute, School of Computing and MathematicsUniversity of UlsterNorthern IrelandUK

Personalised recommendations