Behavioral sciences (Basel, Switzerland)

Using walking patterns and machine learning to identify sensation seeking

Updated

Abstract

The SMO Regression model achieved a performance score of 0.60 in assessing traits using gait analysis.

  • An innovative digital phenotyping approach combines computational gait analysis with machine learning to objectively assess sensation-seeking traits.
  • Data from 233 healthy adults were analyzed, capturing natural gait sequences and self-reported sensation-seeking levels.
  • Twenty-five skeletal keypoints were extracted from gait data, which were then transformed into a hip-centered coordinate system.
  • Three machine learning models were developed and compared, with the SMO Regression model outperforming others in accuracy.
  • The study presents preliminary evidence for using gait patterns as valid psychological indicators through advanced computational techniques.

Simplified

Key numbers

0.60
Gait Model Performance Correlation
Correlation coefficient for SMO Regression model performance.
233
Participant Count
Total number of healthy adult participants in the study.

Full Text

What this is

  • This research explores , a trait linked to mental health issues and risky behaviors.
  • It introduces a digital phenotyping method using gait analysis and machine learning to quantify traits.
  • The study involves 233 healthy adults and compares gait data to self-reported scores.

Essence

  • Gait characteristics can serve as reliable indicators of traits, with machine learning models effectively predicting these traits from gait data.

Key takeaways

  • Gait analysis effectively captures traits, showing significant correlations with self-reported scores. The study demonstrates that gait patterns reflect psychological constructs, suggesting a novel assessment method.
  • The SMO Regression model outperformed other machine learning approaches, achieving a correlation coefficient of 0.60. This indicates its potential for accurately measuring traits based on gait data.
  • The integration of gait analysis and machine learning offers a non-invasive, real-time assessment of . This method could enhance psychological evaluations beyond traditional self-report measures.

Caveats

  • The sample size of 233 and its demographic homogeneity limit the generalizability of the findings. Future studies should validate the model across diverse populations.
  • The study's correlational design does not allow for causal inferences. Longitudinal research is needed to explore the relationships between gait patterns and traits.

Definitions

  • Sensation Seeking: The pursuit of varied, novel, and intense experiences, often involving risks.

Simplified

Funding

Competing interests

The authors declare no conflicts of interest.
PubMed

What Lands in Your Inbox Each Week:

  • 📚7 fresh studies
  • 📝plain-language summaries
  • direct links to original studies
  • 🏅top journal indicators
  • 📅weekly delivery
  • 🧘‍♂️always free