An Ideal Regional Learning Model for Disability Prediction using Provided Input Samples

One type of specific learning disability that might cause academic dissatisfaction is dyslexia. Detection is crucial yet challenging, particularly for languages with simple orthographies. Finding dyslexia remains a major goal for research efforts across many fields, as seen by numerous studies published in numerous scientific journals. These activities, like classifying medical datasets, benefit greatly from deep learning (DL) and other statistical techniques. To detect dyslexia, we have integrated fundamental DL algorithms into this paper. However, a significant amount of manual labor is still required to properly classify the large dataset utilized for training. Recent research has demonstrated that DL models are useful in reducing the workload associated with feature engineering. However, applying the Regional Convolutional Network Model (RCNM) directly to the classification issue does not result in a statistically significant gain in performance. We offer an efficient learning model built on genetic algorithms for better DL prediction. In the initial step of the model, dyslexia features are extracted, and then DL algorithms are used. Next, we merge many DL models by choosing the best weight using an adaptive genetic method. To demonstrate that our suggested method greatly improves accuracy, we ran experiments on the Dyslexia dataset.