Under the current big data background, the training mode of radio and television director technology is obsolete, and the technical means do not meet the needs of modern development. In this article, a self-adaptive multivariate data statistical model of radio and television directors based on the least squares support vector machine is proposed, which combines the students’ views with the diversified teaching methods and teaching contents needed by university teachers in the process of vocational education and television education. This article applies the technology integration degree measurement, market integration degree measurement, business integration degree measurement, and integration degree comprehensive analysis to analyze the data of major video websites and major radio and television media. It is found that the market share of major radio and television media is increasing, and the number of broadcasts of major video online stores is also excellent.