An Innovative Study of Traditional Music Pedagogy Based on Time Series Analysis

This paper proposes an innovative approach to traditional music teaching based on time series analysis. Through the music online learning platform, learners’ music learning behavior data is collected. Using the time series analysis algorithm (ARMA), the changing law of learners’ learning behavior is fitted, and a relevant mathematical model is established. By estimating the model parameters to derive the learning behavior characteristics, trends, and inherent change rules, we can predict the future development of the learners’ music learning effect. In the music online learning platform, the maximum number of logins is 409 times, the longest time for watching videos is 447min, the maximum number of comments is 148 times, and the highest time for browsing forums is 212min. through the behavioral characteristics of the learners, the learners are classified as active participants, content consumers, social learners, and occasional browsers. The algorithm in this paper has better performance in predicting the learning effect on different cluster categories, and the evaluation indexes are all over 0.9. And the algorithm predicts that the average learning effect of the learners in cluster category 4 is only 59.81, which needs to be improved in the learning program. According to the results of time series analysis, this paper proposes four music learning programs. Time series analysis can accurately predict the learning effect of learners in music teaching and formulate learning programs. The teaching effect is better than the traditional teaching method.