Modeling Internet Addiction Severity Using Deep Learning: Effects of Impulsivity, Emotional Dysregulation, Reward Sensitivity, and Fear of Missing Out

Objective: The present study aimed to model internet addiction severity using deep learning techniques by examining the combined effects of impulsivity, emotional dysregulation, reward sensitivity, and fear of missing out.Methods and Materials: This quantitative cross-sectional study was conducted on ۴۱۲ university students from Mexico selected through stratified convenience sampling. Participants completed standardized self-report measures including the Internet Addiction Test (IAT), Barratt Impulsiveness Scale (BIS-۱۱), Difficulties in Emotion Regulation Scale (DERS), Behavioral Activation System (BAS) scale, and Fear of Missing Out (FoMO) scale. Data were collected via an online survey platform. After preprocessing, including handling missing values and standardization, a deep neural network model was developed to predict internet addiction severity. The dataset was divided into training (۷۰%), validation (۱۵%), and test (۱۵%) sets. Model performance was evaluated using mean squared error (MSE), root mean squared error (RMSE), and coefficient of determination (R²), and hyperparameters were optimized using grid search techniques.Findings: The deep learning model demonstrated strong predictive performance, with R² values of ۰.۷۲, ۰.۶۹, and ۰.۶۸ in the training, validation, and test sets, respectively, indicating high explanatory power and generalizability. All predictor variables showed significant positive associations with internet addiction severity, with fear of missing out emerging as the strongest predictor, followed by emotional dysregulation, impulsivity, and reward sensitivity. Shapley value analysis confirmed the relative importance of these variables, highlighting a multidimensional pattern of influence. The model exhibited low error rates across datasets, supporting its robustness and stability in predicting internet addiction severity.Conclusion: The findings indicate that internet addiction severity is best understood as a multidimensional construct influenced by interrelated psychological factors, with fear of missing out and emotional dysregulation playing central roles. The application of deep learning provides a powerful and accurate approach for modeling complex behavioral phenomena, offering both theoretical insights and practical implications for prevention and intervention strategies. Objective: The present study aimed to model internet addiction severity using deep learning techniques by examining the combined effects of impulsivity, emotional dysregulation, reward sensitivity, and fear of missing out. Methods and Materials: This quantitative cross-sectional study was conducted on ۴۱۲ university students from Mexico selected through stratified convenience sampling. Participants completed standardized self-report measures including the Internet Addiction Test (IAT), Barratt Impulsiveness Scale (BIS-۱۱), Difficulties in Emotion Regulation Scale (DERS), Behavioral Activation System (BAS) scale, and Fear of Missing Out (FoMO) scale. Data were collected via an online survey platform. After preprocessing, including handling missing values and standardization, a deep neural network model was developed to predict internet addiction severity. The dataset was divided into training (۷۰%), validation (۱۵%), and test (۱۵%) sets. Model performance was evaluated using mean squared error (MSE), root mean squared error (RMSE), and coefficient of determination (R²), and hyperparameters were optimized using grid search techniques. Findings: The deep learning model demonstrated strong predictive performance, with R² values of ۰.۷۲, ۰.۶۹, and ۰.۶۸ in the training, validation, and test sets, respectively, indicating high explanatory power and generalizability. All predictor variables showed significant positive associations with internet addiction severity, with fear of missing out emerging as the strongest predictor, followed by emotional dysregulation, impulsivity, and reward sensitivity. Shapley value analysis confirmed the relative importance of these variables, highlighting a multidimensional pattern of influence. The model exhibited low error rates across datasets, supporting its robustness and stability in predicting internet addiction severity. Conclusion: The findings indicate that internet addiction severity is best understood as a multidimensional construct influenced by interrelated psychological factors, with fear of missing out and emotional dysregulation playing central roles. The application of deep learning provides a powerful and accurate approach for modeling complex behavioral phenomena, offering both theoretical insights and practical implications for prevention and intervention strategies.

Keywords:

Internet addiction , deep learning , impulsivity , emotional dysregulation , reward sensitivity , fear of missing out , predictive modeling