Modelo de avaliação de risco em campanhas kickstarter utilizando machine learning

Crowdfunding has become a virtual phenomenon because it is a business model that aims to raise funds for collective projects of various kinds. Risk Assessment based on the characteristics of such projects is an important step in minimizing financial losses. In this sense, the objective of this work is to present a model for risk assessment in these projects using Statistical and Machine Learning (ML) techniques, considering the stages of preparation, processing and analysis of the results. For the development of the work the WEKA tool and the Python programming language with ML specific modules were used. The research showed that the use of ML techniques was efficient, obtaining an accuracy of 77%, higher to the 76% of the model proposed by Etter, Grossglauser e Thiran (2013) and can therefore be used as a method in Risk Assessment in Kickstarter Campaigns, providing conditions for investors to mitigate the risks associated with projects of this nature by analyzing graphs and numerical data.