LEVERAGING MACHINE LEARNING FOR A COMPREHENSIVE ANALYSIS OF AUTISM SPECTRUM DISORDER: A SYNTHESIS OF SCHOLARLY INSIGHTS

Autism Spectrum Disorder (ASD) is a complex developmental condition that is defined by difficulties in social interaction, repetitive behaviors, and increased sensitivity to sensory stimuli. Early and accurate diagnosis is crucial for ensuring effective intervention and support. Recent advancements in machine learning have demonstrated promising potential in improving ASD diagnosis using a variety of techniques. This paper reviews recent research on how machine learning is applied to ASD diagnosis, emphasizing the effectiveness of different classification models, feature selection approaches, and their impact on diagnostic accuracy. Key methods include Multilayer Perceptron (MLP) classifiers, which have achieved up to 100% accuracy, along with other advanced AI models that have shown diagnostic accuracies of 98.17%. Additional techniques, such as the Weighted C4.5 Algorithm (WCBA) and Decision Tree Classifiers, have proven successful in identifying critical features and enhancing early detection. Combining feature scaling with various machine learning algorithms has further improved classification performance across diverse age groups. These developments underscore the potential of machine learning to enhance ASD diagnosis, leading to more precise and earlier identification, which is essential for effective intervention and management.