... (combining recurrent neural network and LSTM layers) according to the LSTM structure of the UniBiCLSTM model. This research used the trained models as pretrained models for transfer learning. The authors experimentally demonstrate that the proposed CMTL-UniBiCL-RL model achieved faster and better scenarios...

... by a deep learning model called recurrent neural network (RNN). Finally, the experimental analysis reveals that the proposed QICO-based feature selection model outperforms the other heuristic-based feature selection and optimized RNN outperforms the other machine learning methods. Findings The proposed...

... methodology is to use text-mining package, consisting of the combination of hierarchical recurrent neural networks and Hopfield-like neural network equipped with Laplacian-based graph visualization to visualize the consumers' browsing patterns. Based on the observed interactivity between webpages, consumers...