The New Gated Recurrent Unit Deep Learning System based on GRU and CNN

Abdelhamid. ATASSI and Ikram EL AZAMI

Our paper presents a new deep learning system based on the comparaison of two Systems: the Convolutional Neural Network and the Gated Recurrent Unit, with the addition of two Adadelta and Adam methods to our New Gated Recurrent Unit system, without forgetting to assign a specific values to achieve good results, and the application of this on an IMDb reference database and on a database of film comments. Also, in our future work, the application will be on Arabic tweets to know more about the system's performance.

Volume 11 | Issue 5

Pages: 263-268