Impact of Hybridization of Deep Learning Models for Temporal Data Learning

Various deep learning models work for information retrieval through the learning process. The deep learning models train the machine by learning the data through the neural network architecture. In the learning process, the model architecture affects the performance of the model. The deep hybrid models has given by various researchers in different domains. The hybridization possesses the computational complexity in the models. This paper presents the two ways for hybridization (using convnets and recurrent networks) to analyze the impact on the performance of the hybrid model with less complexity. The experimental results demonstrate the importance of hybridization-based on convnets and recurrent networks. Recurrent-based hybridization performs better in regression while convnets based hybridization in classification on the temporal dataset.