Abstract: Under the Internet information network environment, College English teaching mode is faced with a new choice and transformation. SPOC and STEAM classes have multiple advantages, which can make up for the lack of a single teaching model and provide new ideas for teaching reform. It is a practical application of statistical method to optimize the model of artificial neural network by machine learning method of statistics. The application of mathematical statistics can solve some related problems of artificial perception. Therefore, artificial neural network has the same simple decision ability and judgment ability as human beings. In this paper, the authors analyze the image invariant features and SVM algorithms application in college English education platform. The results show that this method has a positive effect on learners’ English proficiency and learning effect. Teachers also avoid paying a lot of labor, which is very beneficial to the implementation of innovative teaching. However, compared with the traditional teaching, the phenomenon of student achievement differentiation is very serious, and teaching is facing great pressure. Therefore, improving students’ autonomous learning ability and teachers’ information literacy is still very helpful to improve the teaching effect.
Keywords: Artificial neural network, deep learning, automatic recognition, English teaching
