Machine learning-based classifying of risk-takers and risk-aversive individuals using resting-state EEG data: A pilot...

Area of Science:

• Neuroscience
• Cognitive Science
• Machine Learning

Background:

• Decision-making is crucial in management, with growing interest in identifying risk-taking versus risk-averse personality traits.
• Existing research on decision-making signals and brain activity lacks intelligent, brain-based prediction methods for manager risk profiles.
• Distinguishing between risk-taking and risk-averse managers remains a challenge in organizational psychology and neuroscience.

Purpose of the Study:

• To propose and validate an electroencephalogram (EEG)-based intelligent system for distinguishing between risk-taking and risk-averse managers.
• To investigate the efficacy of machine learning algorithms in classifying manager risk profiles using neurophysiological data.
• To explore the potential of brain-based techniques in understanding and predicting managerial decision-making tendencies.

Main Methods:

• Collected resting-state EEG signals from 30 managers.
• Applied wavelet transform for time-frequency analysis to extract statistical features from EEG data.
• Utilized a two-step statistical wrapper algorithm for feature selection, followed by a support vector machine classifier.

Main Results:

• The developed system achieved 74.42% accuracy, 76.16% sensitivity, 72.32% specificity, and 75% F1-measure in distinguishing between risk-taking and risk-averse managers.
• Machine learning models successfully differentiated manager types using features from the alpha frequency band within a 10-second analysis window.
• Demonstrated significant intersubject predictive performance, highlighting the robustness of the EEG-based approach.

Conclusions:

• Intelligent, machine learning-based systems show significant potential for distinguishing between risk-taking and risk-averse managers using biological signals.
• EEG signal analysis, particularly features from the alpha frequency band, can serve as a reliable biomarker for managerial risk propensity.
• This research opens avenues for developing objective, brain-based tools to assess and understand individual differences in managerial decision-making under risk.