Construction of Hybrid Deep Learning Model for Predicting Children Behavior based on their Emotional Reaction

Emotion prediction, the sub-domain of sentiment analysis helps to analyze the emotion. Recently, the prediction of children's behavior based on their present emotional activities is remaining as a challenging task. Henceforth, the deep learning algorithms are used to support and assist in the process of children's behavior prediction by considering the emotional features with a good accuracy rate. Besides, this article presents the prediction of children's behavior based on their emotion with the deep learning classifiers method. To analyze the performance, decision tree and naïve Bayes probability model are compared. Totally, 35 sample emotions are considered in the prediction process of deep learning classifier with a probability model. Furthermore, the hybrid emotions are incorporated in the proposed dataset. The comparison between both the decision tree and the Naïve Bayes method has been performed to predict the children's emotions after the classification. Based on the probability model of naïve Bayes method and decision tree, naïve Bayes method provides good results in terms of recognition rate and prediction accuracy when compared to the decision tree method. Therefore, a fusion of these two algorithms is proposed for predicting the emotions involved in children's behavior. This research article includes the combined algorithm mathematical proof of prediction based on the emotion samples. This article discusses about the future scope of the proposal and the obtained prediction results.