AwGhO_Res_Trans: Adaptive Weighting Grasshopper Optimization-Based Residual Transformer for Autism Detection
Jagadesh Balasubramani, R Surendran
Abstract
In this model, input data is acquired from publicly available datasets and converted into a normalized form using the Min-Max normalization approach. The Pre-processing based on normalization is employed to scale the data within a specific range to reduce complexity. Following normalization, the most relevant features are selected using the Wrapper Feature Selection Model, which iteratively evaluates and selects appropriate features that enhance Autism detection accuracy. Finally, Autism detection is devised using the proposed Adaptive Weighting Grasshopper Optimization Residual Transformer (AwGhO_Res_Trans) model. The designed model demonstrated enhanced outcomes compared to the baseline models in Autism detection.