JACIII Vol.21 No.5 pp. 813-824
doi: 10.20965/jaciii.2017.p0813


Feature Selection Algorithm Considering Trial and Individual Differences for Machine Learning of Human Activity Recognition

Yuto Omae* and Hirotaka Takahashi**

*Department of Electrical Engineering, National Institute of Technology, Tokyo College
1220-2 Kunugida, Hachioji, Tokyo 193-0942, Japan

**Department of Information and Management Systems Engineering, Nagaoka University of Technology
1603-1 Kamitomioka, Nagaoka, Niigata 940-2188, Japan

January 15, 2017
July 4, 2017
September 20, 2017
feature selection, lower dimensional, machine learning, body motion classification, inertia sensor

In recent years, many studies have been performed on the automatic classification of human body motions based on inertia sensor data using a combination of inertia sensors and machine learning; training data is necessary where sensor data and human body motions correspond to one another. It can be difficult to conduct experiments involving a large number of subjects over an extended time period, because of concern for the fatigue or injury of subjects. Many studies, therefore, allow a small number of subjects to perform repeated body motions subject to classification, to acquire data on which to build training data. Any classifiers constructed using such training data will have some problems associated with generalization errors caused by individual and trial differences. In order to suppress such generalization errors, feature spaces must be obtained that are less likely to generate generalization errors due to individual and trial differences. To obtain such feature spaces, we require indices to evaluate the likelihood of the feature spaces generating generalization errors due to individual and trial errors. This paper, therefore, aims to devise such evaluation indices from the perspectives. The evaluation indices we propose in this paper can be obtained by first constructing acquired data probability distributions that represent individual and trial differences, and then using such probability distributions to calculate any risks of generating generalization errors. We have verified the effectiveness of the proposed evaluation method by applying it to sensor data for butterfly and breaststroke swimming. For the purpose of comparison, we have also applied a few available existing evaluation methods. We have constructed classifiers for butterfly and breaststroke swimming by applying a support vector machine to the feature spaces obtained by the proposed and existing methods. Based on the accuracy verification we conducted with test data, we found that the proposed method produced significantly higher F-measure than the existing methods. This proves that the use of the proposed evaluation indices enables us to obtain a feature space that is less likely to generate generalization errors due to individual and trial differences.

  1. [1] A. M. Khan, Y. K. Lee, S. Y. Lee, and T. S. Kim, “A Triaxial Accelerometer-Based Physical-Activity Recognition via Augmented-Signal Features and a Hierarchical Recognizer,” IEEE Tran. on Information Technology in Biomedicine, Vol.14, No.5, pp. 1166-1172, 2010.
  2. [2] J. Lester, T. Choudhury, and G. Borriello, “A Practical Approach to Recognizing Physical Activities,” Int. Conf. on Pervasive Computing, pp. 1-16, 2006.
  3. [3] Z. He and L. Jin, “Activity Recognition from Acceleration Data Based on Discrete Consine Transform and SVM,” IEEE Int. Conf. on Systems, Man and Cybernetics, pp. 5041-5044, 2009.
  4. [4] J. A. Ward, P. Lukowicz, G. Troster, and T. E. Starner, “Activity Recognition of Assembly Tasks Using Body-Worn Microphones and Accelerometers,” IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol.28, No.10, pp. 1553-1567, 2006.
  5. [5] P. Siirtola, P. Laurinen, J. Roning, and H. Kinnunen, “Efficient Accelerometer-Based Swimming Exercise Tracking,” Computational Intelligence and Data Mining (CIDM), 2011 IEEE Symp., pp. 156-161, 2011.
  6. [6] Y. Kon, Y. Omae, K. Sakai, H. Takahashi, T. Akiduki, C. Miyaji, Y. Sakurai, N. Ezaki, and K. Nakai, “Toward Classification of Swimming Style by Using Underwater Wireless Accelerometer Data,” ACM Int. Symp. on Wearable Computers, pp. 85-88, 2015.
  7. [7] T. Hastie, R. Tibshirani, J. Friedman, M. Sugiyama, T. Ide, T. Kamishima, T. Kurita, and E. Maeda, “The Elements of Statistical Learning: Data Mining, Inference, and Prediction,” Kyoritsu Publisher, 2014.
  8. [8] L. Breiman, “Random Forests,” Machine Learning, Vol.45, No.1, pp. 5-32, 2001.
  9. [9] M. Robnik-Sikonja and I. Kononenko, “Theoretical and Empirical Analysis of ReliefF and RReliefF,” Machine Learning, Vol.53, No.1-2, pp. 23-69, 2003.
  10. [10] X. W. Chen and C. J. Jong, “Minimum Reference Set Based Feature Selection for Small Sample Classifications,” The 24th Int. Conf. on Machine Learning, pp. 153-160, 2007.
  11. [11] S. Dudoit, J. Fridlyand, and T. P. Speed, “Comparison of Discrimination Methods for the Classification of Tumors Using Gene Expression Data,” J. of the American Statistical Association, Vol.97, No.457, pp. 77-87, 2002.
  12. [12] J. Ye, T. Li, T. Xiong, and R. Janardan, “Using Uncorrelated Discriminant Analysis for Tissue Classification with Gene Expression Data,” IEEE Trans. on Computational Biology and Bioinformatics, Vol.1, No.4, pp. 181-190, 2004.
  13. [13] C. A. Ronao and S. B. Cho, “Human Activity Recognition Using Smartphone Sensors with Two-Stage Continuous Hidden Markov Models,” 10th Int. Conf. on Natural Computation, pp. 681-686, 2014.
  14. [14] Y. Omae, Y. Kon, M. Kobayashi, K. Sakai, A. Shionoya, H. Takahashi, T. Akiduki, K. Nakai, N. Ezaki, Y. Sakurai, and C. Miyaji, “Swimming Style Classification Based on Ensemble Learning and Adaptive Feature Value by Using Inertial Measurement Unit,” J. of Advanced Computational Intelligence and Intelligent Informatics, Vol.21, No.4, pp. 616-631, 2017.
  15. [15] A. Wang, G. Chen, J. Yang, S. Zhao, and C. Y. Chang, “A Comparative Study on Human Activity Recognition Using Inertial Sensors in a Smartphone,” IEEE Sensors J., Vol.16, No.11, pp. 4566-4578, 2016.
  16. [16] R. Akhavian, L. Brito, and A. Behzadan, “Integrated Mobile Sensor Based Activity Recognition of Construction Equipment and Human Crews,” Conf. on Autonomous and Robotic Construction of Infrastructure, 2015.
  17. [17] Y. Kon, Y. Omae, K. Sakai, H. Takahashi, T. Akiduki, C. Miyazi, Y. Sakurai, N. Ezaki, and K. Nakai, “Swimming Style Classification for Developing System of Swimming Performance and Technique Evaluation,” JSME Symp.: Sports engineering and Human Dynamics 2015, A-15, 2015.
  18. [18] Y. Ohgi, K. Kaneda, and A. Takakura, “A swimming style prediction using the chest acceleration,” Symp. on Sports and Human Dynamics 2012, pp. 98-103, 2012.
  19. [19] W. Choi, J. Oh, T. Park, S. Kang, M. Moon, U. Lee, I. Hwang, and J. Song, “Mobydick: An Interactive Multi-Swimmer Exergame,” The 12th ACM Conf. on Embedded Network Sensor Systems, pp. 76-90, 2014.
  20. [20] U. Jensen, F. Prade, and B. M. Eskofier, “Classification of Kinematic Swimming Data with Emphasis on Resource Consumption,” Body Sensor Networks (BSN), 2013 IEEE Int. Conf., pp. 1-5, 2013.
  21. [21] L. Bao and S. S. Intille, “Activity Recognition from User-Annotated Acceleration Data,” Pervasive Computing, pp. 1-17, 2004.
  22. [22] N. Ravi, N. Dandekar, P. Mysore, and M. L. Littman, “Activity Recognition from Accelerometer Data,” AAAI, Vol.5, pp. 1541-1546, 2005.
  23. [23] M. Kobayashi, Y. Omae, Y. Kon, K. Sakai, A. Shionoya, H. Takahashi, Y. Sakurai, C. Miyaji, K. Nakai, K. Nakai, N. Ezaki, and T. Akiduki, “Analysis of Stroke Duration for Swimming Motion Coaching System by Using a Sensor Device,” The Robotics and Mechatronics Conf. 2016 (ROBOMECH2016), 2A1-11b5-1-4, 2016.
  24. [24] Sports Sensing Co., Ltd., “9 Axis Waterproof Type Wireless Motion Sensor,”, [accessed Feb. 9, 2016].
  25. [25] Sony, Digital Video Camera HDR-CX720V,, [accessed Feb. 17, 2016].

*This site is desgined based on HTML5 and CSS3 for modern browsers, e.g. Chrome, Firefox, Safari, Edge, IE9,10,11, Opera.

Last updated on Oct. 16, 2017