Research Article
Feature Signature Discovery for Autism Detection: An Automated Machine Learning Based Feature Ranking Framework
Table 1
Summary of research findings from the literature on ASD classification using machine learning models.
| Year | Data | Model | Features | Tool | Accuracy (%) | Evaluation |
| 2019 | Adult, adolescent, child—UCI [9] | SVM and RF | All | Matlab | 100 | 90%–10% train-test | 2021 | Toddler data—Kaggle [24] | Neural network | Not specified | R-studio | 99 | Test data | 2021 | Adult, adolescent, child—UCI, and toddler [7] | Stochastic gradient descent, RF, AdaBoost | Not specified | Not mentioned | ∼97 | Not mentioned | 2022 | Adult, adolescent, child—UCI [8] | Voting meta classifier | Not specified | Flask web app | ∼91–97 | Stratified 10-fold cross-validation |
|
|
SVM-support vector machine; RF-random forest. |