Treffer: Prediction model for the success rate of college student entrepreneurship projects based on deep learning.

The rise in innovation and entrepreneurship in higher education has sparked a surge in student-led startups. Still, many fail due to poor planning, inadequate mentorship, and insufficient resource allocation, making it challenging to accurately predict their success. This research aims to develop a robust and scalable prediction model to estimate the success rate of college entrepreneurship projects using a novel Scalable Tasmanian Devil-driven Adaptive Deep Neural Network (STD-ADNN) architecture. A dataset of 2100 labelled projects from 40 institutions spanning 5 years was compiled from national student entrepreneurship competitions, incubation centers, and innovation funding programs. The dataset includes project type, team details, innovation level, mentorship availability, funding, and market-readiness indicators. The research utilized Min–Max normalization and label encoding techniques to scale numerical features like funding amount and team size, as well as convert categorical features like project domain and institutional category into machine-readable formats. Principal Component Analysis (PCA) feature extraction technique is used to identify and retain the most significant features, enhancing model accuracy and reducing computational complexity. The STD-ADNN model, inspired by the Tasmanian devil's adaptive foraging behavior, optimizes resource selection and prioritization. It adjusts input feature weights based on predictive value, reinforcing informative attributes and suppressing less relevant ones during training. This adaptive weighting strategy improves convergence speed, model generalization, and predictive accuracy, enhancing learning focus and model generalization. The model, implemented using Python, achieved an accuracy of 0.978 and outperformed traditional classifier models. This predictive framework provides institutions and incubators with a powerful data-driven tool to identify high-potential student projects and improve entrepreneurial outcomes in academic settings. [ABSTRACT FROM AUTHOR]

Copyright of Discover Artificial Intelligence is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)