This paper proposes forecasting loss aversion bias using hybrid machine learning models. It uniquely incorporates reaction time, psychological factors (self-confidence, hopelessness, loss aversion), and personality traits (financial literacy, socio-demographics) as features. Random Forest emerged as the superior algorithm. A new phenomenon, "irresistible impulse of gambling," was identified where decision-makers minimize decision duration as the positive spread ratio converges to the default loss aversion level.

Ömür Saltık, Wasim ul Rehman, Rıdvan Söyü, Süleyman Değirmen, Ahmet Şengönül