Research on Analyzing the Competitive Situation of College Students' Innovation and Entrepreneurship Market Based on the Algorithm of Multilayer Perceptual Machine

Abstract

In recent years, along with more and more employment problems, the state has formulated a series of policies to support college students' innovation and entrepreneurship, and all colleges and universities are actively exploring the competitive market dynamics. This paper starts from the current situation of college students' innovation and entrepreneurship to understand its main influencing factors. The traditional TF-IDF algorithm is then used to extract text features, and it is found that there are defects in the algorithm, for which the importance value is introduced on the basis of the original algorithm. The extracted text feature vector of college students' innovation and entrepreneurship market competition situation is used as the input layer of the multilayer perceptron algorithm, and the output layer is the prediction value of the market competition situation, and the BP neural algorithm is used for training, which finally completes the task of constructing the prediction model of the college students' innovation and entrepreneurship market competition situation, and empirically analyzes the model. In terms of AUC area, the performance of BP algorithm (index difference: 0.1052) is better than that of random forest algorithm (index difference: 0.1194), which proves that this paper's algorithm is prioritized in the analysis of competitive situation of college students' innovative entrepreneurship market. In addition, this paper's algorithm detects that the number of employment in the tertiary industry increases by 1,479,600 people every year, and the corresponding number of employment in the primary industry and the secondary industry shows a decreasing trend, which quantitatively reflects the current competitive situation in the market of college students' innovation and entrepreneurship.