Lattice-based cryptography, as an active branch of post-quantum cryptography (PQC), has received extensive attention from side-channel analysis (SCA) researchers in recent years. The TALE is a candidate for China's post-quantum project, which aims to standardize cryptographic systems from attacks from quantum and classical computers. Although TALE relies on the theory of quantum-lattice resistance, practical implementations still have the vulnerability of side-channel analysis (SCA). In side-channel analysis (SCA), template attack is considered to be the most effective analysis method, and traditional template analysis are now gradually being replaced by machine learning-based template attack due to low computational efficiency. In this paper, the TALE is analyzed for vulnerability, feature extraction is performed for the leakage information of vulnerable points, and a template attack method based on multilayer perceptron (MLP) is used to attack vulnerable points.