Personalized Teaching Strategy Suggestion for Toddlers with ASD: A Machine Learning and Deep Learning Approach

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorders which have impacts particularly on social interaction, communication, cognitive development, and adaptive learning. Since every infant with ASD has different autism screening characteristics and skill gaps, it is necessary to choose the best teaching strategy in order to meet their educational demands. In order to fulfill the developmental needs of each toddler with ASD, we want to determine the best teaching strategies that are well suited according to their learning requirements. To provide a suitable teaching strategy for toddlers, this study introduces an approach of Machine Learning (ML) and Deep Learning (DL) which analyzes the distinctive characteristics of autism. We have used here Analysis of Variance (ANOVA) feature selection technique to see how each feature influences the target variable. We have applied Decision Tree (DT), Random Forest Classifier (RFC), Feedforward Neural Network (FFNN), Deep Neural Network (DNN) for classifying the most appropriate teaching strategy. The FFNN and DNN performs outstanding with 100% accuracy.