Research on Optimising Personalised Teaching Models in University Piano Courses Using Reinforcement Learning Algorithms

University music and art teaching is an important link to improve students ‘literary literacy, enhance students’ cultural confidence, and improve students’ comprehensive ability. And for the university music traditional education mode exists a low degree of feedback, teacher-student interaction is difficult, teachers are difficult to accurately and quickly grasp the students’ movements and other problems, which is not conducive to the improvement of teaching quality. Personalised teaching mode can be a good solution to the above problems, but due to its limitations, it has not been used on a large scale in the past. With the rapid implementation of informatisation and digitalisation in the information age, it provides technical support for the large-scale popularisation of personalised education. The purpose of this paper is to explore the optimisation method of the machine algorithm model for personalised teaching mode of university music classroom in information technology environment. We introduced deep learning technology and constructed a personalised teaching feedback model based on dynamic grouping, automatic information collection system and graph convolutional neural network fusion. And we designed a personalised teaching model for university piano personalised teaching supported by personalised recommendation system. Finally, the performance and accuracy of the constructed model are verified by comparison. The results show that the personalised teaching model constructed in this paper has achieved good application results in piano teaching, with students’ satisfaction reaching nearly 100% in a semester’s learning time, while the unique weight allocation table further ensures that the model is always consistent with the goals of university music education during the optimisation process. The constant feedback and dynamic grouping during the optimisation of the learning algorithms described above allows for a rapidity and subjectivity that is not possible in a static model, which is in line with the educational requirements for student and teacher motivation and autonomy in curriculum design. The model can be used not only in the personalised classroom design of music teaching, but also to optimise the traditional teaching model, which also provides a reference value for educators to implement personalised teaching.