A machine learning approach for analyzing sexual satisfaction based on psychological features

Abstract

The emergence of machine learning techniques has revolutionized various fields, helping to shed light into the complexities of human sexuality and address sexuality-related problems. The present study aimed to classify sexual satisfaction in both women (n = 503) and men (n = 342), who completed a digital survey aimed at Colombian adults based on a snowball sampling. Collected data were analyzed using several supervised learning algorithms where inputs included marital status, sociosexuality, sexual drive, sexual functioning, and personality traits. The results showed that the XGBoost model provided best classification results for sexual satisfaction in women, while the Artificial Neural Networks (ANN) had the best performance in classifying sexual satisfaction in men. In both groups, sexual functioning and sexual drive were the most significant predictors of sexual satisfaction. Traits such as extraversion, narcissism, machiavellianism, and sociosexual behavior had a lesser importance. Lastly, psychopathy emerged as a significant predictor of men's sexual satisfaction, whereas conscientiousness emerged as a significant predictor of women's satisfaction. This study provides a technological tool to classify sexual satisfaction using Machine Learning models, in addition, provide, in terms of entropy, variables with the greatest influence based on the data and predictions.