LiteLSTM architecture based on weights sharing for recurrent neural networks

作者：Nelly Elsayed, Zag ElSayed, Anthony S. Maida · 发表于：International Journal of Computers and Applications · 年份：2025 · DOI：10.1080/1206212x.2025.2499869 · 被引用次数：2 · 研究领域：Neural Networks and Applications、Handwritten Text Recognition Techniques、Time Series Analysis and Forecasting

Long short-term memory (LSTM) is one of the robust recurrent neural network architectures for learning sequential data. However, it requires considerable computational power to learn and implement both software and hardware aspects. This paper proposed a novel LiteLSTM architecture based on reducing the LSTM computation components via the weights sharing concept to reduce the overall architecture computation cost and maintain the architecture performance. The proposed LiteLSTM can be significant for processing large data at different domains where time-consuming is crucial while hardware resources are limited, such as the security of Internet of Things (IoT) devices and medical data processing. The proposed model was evaluated and tested empirically on three different datasets from the computer vision, cybersecurity, and speech emotion recognition domains. The proposed LiteLSTM has comparable accuracy to the other state-of-the-art recurrent architecture while using a smaller computation budget.