Scholarlink

<p>This study investigates the potential of machine learning (ML) for predicting food-service crowdfunding success, addressing a gap in online crowdfunding platform and smart business ecosystem. Drawing on costless signaling theory and insights from crowdfunding literature, five reputational and eleven non-reputational attributes were identified as key determinants. Data from 22,923 Kickstarter projects was analyzed using seven supervised ML algorithms: bootstrap aggregating ensembles, classification tree, logistic regression, random forest, support vector machine, XGBoost, and deep learning. The results indicate that deep learning is the most accurate model for predicting crowdfunding success. Notably, XGBoost also demonstrated strong predictive power, offering a viable alternative to deep learning. This research pioneers the application of supervised ML in predicting food-service crowdfunding success, expanding the scope of ML applications within hospitality studies and introducing novel predictors. The findings provide valuable insights for entrepreneurs seeking funding through crowdfunding platforms and contribute to the understanding of success factors in this domain.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AI-powered predictive modeling for food-service crowdfunding success: An integrated approach with business intelligence and supervised machine learning for smart business ecosystem

  • 期刊论文

  • 发表日期：

    2025年9月3日

查看全文
  • Minwoo Lee,
  • Yoon Koh,
  • Araceli Hernandez,
  • Taehyee Um

摘要

This study investigates the potential of machine learning (ML) for predicting food-service crowdfunding success, addressing a gap in online crowdfunding platform and smart business ecosystem. Drawing on costless signaling theory and insights from crowdfunding literature, five reputational and eleven non-reputational attributes were identified as key determinants. Data from 22,923 Kickstarter projects was analyzed using seven supervised ML algorithms: bootstrap aggregating ensembles, classification tree, logistic regression, random forest, support vector machine, XGBoost, and deep learning. The results indicate that deep learning is the most accurate model for predicting crowdfunding success. Notably, XGBoost also demonstrated strong predictive power, offering a viable alternative to deep learning. This research pioneers the application of supervised ML in predicting food-service crowdfunding success, expanding the scope of ML applications within hospitality studies and introducing novel predictors. The findings provide valuable insights for entrepreneurs seeking funding through crowdfunding platforms and contribute to the understanding of success factors in this domain.