Facial expression recognition is a crucial component of emotion research and a prerequisite for human-machine interface.In general, face detection, feature extraction, and feature classification are part of a facial expression recognition system.Although traditional machine learning techniques have achieved significant success, the majority of them have difficult computational issues and cannot extract complete and abstract information.Deep learning-based techniques can achieve a greater detection accuracy for facial emotions, but they have a high hardware need and require a lot of training data and tuning parameters.In order to address the aforementioned issues, this paper suggests a method that combines features extracted by a convolutional neural network (CNN) with the C4.5 classifier to identify facial expressions. This method not only identifies the deficiencies of manually created features but also avoids the need for a deep learning model with a high hardware configuration.Random forest is also used to address several issues with C4.5classifier so it is based on the concept of ensemble learning, which is a process of combining multiple classifiers to solve a complex problem and to improve the performance of the model.By combining these two methods, the proposed system achieves high-level accuracy and these procedures enable the identification of the facial emotions: anger, disgust, surprise, sadness, fear, happiness, and neutral.Performance of the proposed system is evaluated using JAFFE and CK+ data sets.

Convolutional neural network(CNN),Random Forest(RF),JAFFE,CK