In this paper, we first introduce the transformer-encoder with a self-attention mechanism and propose a Convolutional Neural Network-Transformer Encoder (CNN-TE) framework to model the temporal ...