An Estimation Method of Intellectual Concentration State by Machine Learning of Physiological Indices

  • Kaku KimuraEmail author
  • Shutaro Kunimasa
  • You Kusakabe
  • Hirotake Ishii
  • Hiroshi Shimoda
Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 903)


Although recent information society has improved the value of intellectual work productivity, its objective and quantitative evaluation has not been established. It is suggested that intellectual productivity can be indirectly evaluated by estimating intellectual concentration states when giving cognitive load. In this study, therefore, the authors have focused on physiological indices such as pupil diameter and heart rate which are supposed to be closely related to cognitive load in office work, and an estimation method of intellectual concentration states from the measured indices has been proposed. Multiple patterns of classification learning methods such as Decision Tree, Linear Discrimination, SVM, and KNN were employed as the estimation method. Based on the estimation method, an evaluation experiment was conducted where 31 male university students participated and the measured psychological indices were given to the classification learning estimators.


Intellectual concentration state Machine learning Physiological indices 



This work was supported by JSPS KAKENHI Grant Number JP17H01777.


  1. 1.
    Tryon, W.W.: Pupillometry: a survey of sources of variation. Psychophysiology 12(1), 90–93 (1975)CrossRefGoogle Scholar
  2. 2.
    Jorna, P.G.A.M.: Spectral analysis of heart rate and psychological state: a review of its validity as a workload index. Biol. Psychol. 34(2), 237–257 (1992)CrossRefGoogle Scholar
  3. 3.
    Hess, E.H., Polt, J.M.: Pupil size in relation to mental activity during simple problem-solving. Science 143(3611), 1190–1192 (1964)CrossRefGoogle Scholar
  4. 4.
    Mulder, G., Mulder, L.J.M.: Information processing and cardiovascular control, psychophysiology. Psychophysiology 18(4), 392–402 (1981)CrossRefGoogle Scholar
  5. 5.
    Inc MathWorks: MATLAB. Accessed May 2018
  6. 6.
    Dietterich, T.G., Bakiri, G.: Solving multiclass learning problems via error-correcting output codes. J. Artif. Intell. Res. 2, 263–286 (1995)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Kaku Kimura
    • 1
    Email author
  • Shutaro Kunimasa
    • 1
  • You Kusakabe
    • 1
  • Hirotake Ishii
    • 1
  • Hiroshi Shimoda
    • 1
  1. 1.Graduate School of Energy ScienceKyoto UniversityKyotoJapan

Personalised recommendations