Abstract:
This study analyzes the use of machine learning algorithms to predict Attention Deficit
Hyperactivity Disorder (ADHD) based on a number of feature sets. A dataset from online
containing demographic information, personality characteristics, and medical indicators
was used. Among the classifiers used, the logistic regression model achieved the highest
accuracy of 98.17%. Data preprocessing consisted of cleaning, transforming, and encoding
features, followed by feature selection and exploratory data analysis. The logistic
regression model, trained on the preprocessed dataset, outperformed other classifiers
including random forest, voting, gradient boosting, Gaussian Naive Bayes, and decision
tree models, with an accuracy of 98.17%. The model showed high precision, recall, and
F1-score, indicating that it was effective in distinguishing between people with and without
ADHD. The dataset's attributes were important in identifying key predictors of ADHD,
providing helpful information for clinical decision-making. Ethical concerns about data
privacy and algorithmic bias were addressed all through the study to ensure responsible
implementation. Overall, this study shows the potential of machine learning to improve
ADHD diagnosis and shows the importance of comprehensive data collection and rigorous
model evaluation in healthcare applications.
