Innovation and Entrepreneurship Methods and Ability Enhancement of College Students in Higher Vocational Colleges and Universities in the New Era Based on Bayesian Statistics

Abstract

As higher education thrives and evolves in the contemporary world, a new paradigm for the advancement of additional practical abilities is the combination of entrepreneurship and creative education. This manuscript proposes the use of Bayesian statistics to enhance the ability of college students in Higher Vocational Colleges and Universities (IEAECS-HVCU-NEBS-EPTANN) to innovate and entrepreneurship methods. The first source of the input data is the 2015 dataset from the Kauffman Entrepreneurship Education Inventory Four-Year Colleges (KEEI). Then, the collected data is fed into pre-processing utilizing Distributed Minimum Error Entropy Kalman Filter (DMEEKF). The DMEEKF is used to Cleaning up the data, Integrating the data, data generalization and data transformation. Then the preprocessed data undergoes Signed Cumulative Distribution Transform (SCDT) for feature extraction. SCDT extract statistical features such us entropy, energy, variance, mean and standard deviation. Extracted features are fed to features selection. Here, it selects 16 features by utilizing Fox-inspired Optimization Algorithm (FOA). The Efficient Predefined Time Adaptive Neural Network (EPTANN) is then given the chosen characteristics in order to forecast and categorise as either a course or no course for higher professional colleges' instruction on innovation and firm ownership. Generally speaking, EPTANN does not represent optimisation techniques that may be adjusted to find the best parameters for predicting company ownership and innovation education courses in higher professional institutions. Hence, the Fractional Pelican African Vulture Optimization(FPAVO)is used to optimize EPTANN which accurately classifies the courses in higher professional universities that educate about innovation and company ownership. The proposed IEAECS-HVCU-NEBS-EPTANN approach is implemented in Python. Using performance criteria including accuracy, precision, recall, F1-score, MSE, and ROC, the proposed method's effectiveness was evaluated. The proposed IEAECS-HVCU-NEBS-EPTANN approach contains 29.9%, 28.5% and 26.8% higher accuracy, 27.46%, 25.68% and 18.79% higher F1-Score and15.79%, 18.51% and 24.61% lower MSE compared with existing methods, such as Research on the practice of innovation and entrepreneurship education in universities under the environment of big data (IEEU-EBD-FCM), research on the construction problems of innovation and entrepreneurship education programmes in higher vocational colleges and universities against the backdrop of the digital technology era (IEEP-HVCU-SVM), and research on the function of structural equation model analysis in higher education agglomeration and innovation and entrepreneurship (CSE-HEAIE-SEM).