,
Research Scholar, Department of Computer Science and Engineering, Jain (Deemed-To-Be University), Bengaluru, Karnataka, India
,
Associate Professor, Department of Computer Science and Engineering, Jain (Deemed-To-Be University), Bengaluru, Karnataka, India
,
Associate Professor, Department of Computer Science and Engineering, JAIN (Deemed-to-be University), Bangalore, Karnataka, India
Professor, Department of Computer Science and Engineering, Jain (Deemed-To-Be University), Bengaluru, Karnataka, India
The classification of ASD from neuroimaging and behavioral information has not been easy owing to factors such as high dimensionality, non-linearity, and overfitting, among others. In this research, propose an innovative framework based on QBMs for feature selection and efficient representation for ASD classification. QBMs have been used to leverage quantum-motivated energy landscape search and sparsity-based regularization strategy for selection of discriminative features and exclusion of redundant features. After preprocessing and feature encoding stages, QBM-based framework is applied to learn feature relationships and extract compact set of relevant features. Benchmark ASD data sets consisting of neuroimaging and behavioral features were used for experimental testing. This suggested model manages to get a result of 98.6% accuracy, 97.4% precision, 97.4% recall, and 97.4% of the F1 score. Also, the suggested model is able to reduce the number of features from 18 to only 9. When comparing the results that this framework produces with other traditional machine learning approaches, including K-Nearest Neighbors (KNN), LASSO, Random Forest, and Gradient Boosting Classifier (GBC), it can be said that this model is on par with them in terms of accuracy. Ablation study reveals the role played by the regularization term in reducing features and improving the performance of the framework. This study shows that quantum feature selection technique can be employed for analyzing large high-dimensional biomedical datasets, especially in ASD diagnosis.
This is an open access article distributed under the Creative Commons Attribution Non-Commercial License (CC BY-NC) License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The statements, opinions and data contained in the journal are solely those of the individual authors and contributors and not of the publisher and the editor(s). We stay neutral with regard to jurisdictional claims in published maps and institutional affiliations.
