| Summary: | 碩士 === 國立屏東大學 === 商業自動化與管理學系碩士班 === 106 === The financial industry is seeing rapid introduction of new technologies and new business models that are challenging established practices. Raising funds successfully is crucial for a crowdfunding campaign. There has been intensive research from academics and practitioners regarding models for predicting success of crowdfunding project. Prior academic research has evaluated success using traditional statistics techniques (e.g. discriminant analysis and logistic regression) and early artificial intelligence models (e.g. artificial neural networks). This study used Indiegogo, one of the early public fundraising platforms, as the research object, and collected all projects between 2015 and 2017. After preliminary statistical analysis, four machine learning techniques are applied and compared for prediction performance. The results show random forest has better forecasting performance. The managerial implication of this research is that entrepreneurs can apply the proposed methodology to identify the most influential topical features embedded in project, and improve the chance of raising sufficient funds for their projects.
|
|---|
