This study explores reliable approaches to identifying Attention-Deficit/Hyperactivity Disorder (ADHD), a neurodevelopmental condition impacting various aspects of life. While traditionally diagnosed through subjective clinical evaluation, this work examines the integration of sensory data and machine learning techniques for more objective ADHD detection. Investigating diverse machine learning algorithms, including Logistic Regression (LR), Random Forest (RF), XGBoost (XGB), LightGBM (LGBM), Neural Network (NN), and Support Vector Machine (SVM), the research analyzes both activity and heart rate variability (HRV) data from a dataset of 103 participants. Results indicate comparable performance between activity and HRV data individually, with notable improvement seen in a combined dataset. The SVM model emerges as the top performer, achieving an F1-Score of 0.87 and a Matthews Correlation Coefficient of 0.77. This study underscores the great potential of interdisciplinary collaboration and diverse data resources in advancing ADHD detection through innovative machine learning techniques.