Student Entrepreneurial Intention Prediction: A Machine Learning-Based Entrepreneurial Mindset Dimension Analysis Method

This paper explores the impact of entrepreneurial mindset dimensions on students’ entrepreneurial intentions from an educational technology perspective, using machine learning techniques to analyze key mindset factors such as innovation orientation, risk tolerance, opportunity recognition, self-efficacy, and persistence, along with control variables like age, gender, major, family background, and entrepreneurship education. Models including Linear Regression, Random Forest, XGBoost, and SVR are employed, with predictive performance evaluated through metrics like R², accuracy, and F1 score. The dataset comprises 1000 simulated student records, ensuring realistic distributions. Results show XGBoost achieves the highest R² of 0.36, indicating moderate explanatory power. Random Forest classification accuracy reaches 0.735. Feature importance highlights interactions such as innovation orientation with persistence as key predictors. Intervention simulations demonstrate a 12% increase in intentions after enhancing key dimensions, supporting targeted educational strategies for online platforms. These findings contribute to entrepreneurial education by identifying influential mindset dimensions and proposing ML-based interventions, with implications for policy and practice. Limitations include simulated data, suggesting future validation with real surveys.