This study presents a decision support system for the early diagnosis of digital addiction observed in Generation Z individuals. Born in the digital age, Generation Z actively uses the internet and social media as a means of social interaction, which leads to digital addiction. At this stage, when identity formation has just begun, digital addiction causes negative consequences such as social isolation, academic failure, anxiety, and depression. The aim of the study is to model the levels of digital addiction using machine learning methods by observing the effects of smartphone usage habits among students aged 12 to 21 in this context. This model has been created by observing daily phone usage statistics. The publicly available data set used in the study has been collected through structured surveys. The dataset shows that screen time is 4.5 hours across all age groups. A relationship between screen time exceeding eight hours and addiction has also been observed. Four different classification algorithms (Logistic Regression, Gradient Boosting, Neural Net (MLP), XGBoost) have been used in the study. Among the models, the Logistic Regression model showed the highest accuracy in classification performance. Compared to similar studies in literature, the machine learning approach has higher prediction success in classifying the level of digital addiction. This study, which applies a data-driven analytical approach to the problem of digital addiction in early ages when identity formation is just beginning, emphasizes the significance of developing early diagnosis and intervention strategies.