A Stochastic Gradient Descent Support Vector Classification (SGD-SVC) Empowered Framework for Precise, Early Autism Detection in Toddlers
Year: 2025
Type: preprint
Abstract: The ability to accurately and early recognize the autism spectrum disorder (ASD) is a critical factor in enabling effective early intervention strategies. This study proposes a novel framework that conceptualizes a synergistic combination of Stochastic Gradient Descent (SGD) and Support Vector Classification (SVC). The framework applies SGD to iteratively refine feature weights, effectively minimizing classification error and identifying the most salient indicators of ASD from a high-dimensional input space. The optimized features are fed into an SVC model for robust pattern recognition. The proposed model's robust performance was verified through extensive experiments on existing benchmark ASD datasets, which demonstrated consistent superiority in classification accuracy, sensitivity, and specificity when compared to existing classifications based on methods derived from traditional machine learning approaches. These results demonstrate that the GD-SVC framework could offer a reliable tool to help clinicians diagnose children with autism early, ultimately meeting the clinical requirement for increased individualized treatment planning and support.
Authors: Ritika Wason, Parul Arora, Akshay Sobti, M N Hoda
Institution: Bharati Vidyapeeth (Deemed to be University)
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: gold
OpenAlex ID: W4414977966
Cites:
Cited by:
Related to: