Chronological pelican remora optimization-enabled deep learning for detection of autism spectrum disorder

The foremost goal of this investigation is to construct a productive paradigm for ASD detection wielding devised Chronological Pelican Remora Optimization Algorithm (CPROA). Initially, median filtering is used to eliminate distortions during the pre-processing stage, and the selected portion is extricated using ROI extraction. After that, the nub-region is extracted based on functional connectivity using the proposed Pelican Remora Optimization (PRO), which is the combination of Pelican Optimization Algorithm (POA) and Remora Optimization Algorithm (ROA). The final step is to classify ASD into normal and abnormal conditions by exploiting Deep Convolutional Neural Network, where the classifier is trained by using CPROA. The newly introduced CPROA is obtained by the amalgamation of chronological principle with POA and ROA. The designed model resulted in high performance with high accuracy of 0.952, recall of 0.958, and F1-score of 0.963.