WH2014 Session: Machine learning-based analysis of smartphone usage pattern and

WLSACONVERGENCE SUMMIT

MACHINE LEARNING-BASED ANALYSIS OF SMARTPHONE USAGE PATTERN AND ITS ASSOCIATION WITH NEGATIVE HUMAN EMOTION

GALEN CHIN-LUN HUNG

What you do on the smartphone predicts how

you feel.

Our model predicts 3 negative emotions based on the smartphone

usage pattern in the past two hours, with an average accuracy rate of

83.2%

Pleasure AnxietyStress

No feature selectionSelected model

DEMO & ABSTRACT PRESENTATION SESSION #11.4 RESEARCH ABSTRACT

MACHINE LEARNING-BASED ANALYSIS OF SMARTPHONE USAGE PATTERN AND ITS ASSOCIATION WITH NEGATIVE HUMAN

EMOTIONGALEN CHIN-LUN HUNG; PEI-CHING YANG; CHIA-CHI CHANG;

JUNG-HSIEN CHIANG

This model has a potential for real-time treatment by envisioning negative emotions

and providing preventive interventions before the users actually 'feel bad’

WLSACONVERGENCE SUMMIT

www.wirelesshealth2014.org