"Non-Invasive Autism Detection Using MLP Neural Network"

"Early detection of Autism Spectrum Disorder (ASD) is essential for timely intervention. This study presents a non-invasive autism detection system using a Multi-Layer Perceptron (MLP) neural network. Physiological and behavioral data are preprocessed and encoded to provide high-quality inputs for the model. The system achieves high accuracy and robust generalization, as validated by train-test evaluation, confusion matrix, and ROC analysis. The proposed approach demonstrates the potential of machine learning for effective, non-invasive autism screening, offering a practical tool for clinicians and caregivers."