Microcontroller based Automatic Sun Tracking Solar Panel

To learn a hierarchal representation of data, deep learning techniques can be used that use multiple processing layers, and produce state of art results. Many models and methods are designed in deep learning for classification in natural language processing (NLP). Various classification algorithms have been used for Arabic documents classification, but they have two problems High dimensional feature representation and the low accuracy of the classification. In this work, an important experiment is made by using deep related models and methods for classifying Arabic text also compare our model with various models. Also to forward a full understand, present and future of deep learning in Arabic text classification and have obtained encouraging results.