providing a dynamic investment model for financing knowledge-based companies with a data mining approach

The development of computer technologies and automated learning techniques can make decision-making easier and more efficient. in the field of machine learning, where computers always make decisions or propose suggestions for proper decision-making, there exist many decision-making techniques such as decision trees, neural networks, etc. flexibility and comprehensibility are one of the advantages of the decision tree model. the decision tree can provide the possible options, goals, financial profit, and information needed for an investment for the managers better than any other tool. the decision tree is one of the most applicable data mining algorithms. on the other hand, crowdfunding in knowledge-based companies is a new financial phenomenon in online financing of innovative projects and knowledgebased businesses that reduces financing costs and problems in addition to changing the nature of the investment. there are four types of crowdfunding in knowledge-based companies namely donation-based, equity-based, lendingbased, and reward-based. reward-based crowdfunding can be considered the most publicly familiar crowdfunding model, where backers will actively participate in the product development process along with investment. low-cost crowdfunding websites act in the projects as an online mediatory between the initiators and the sponsors. therefore, the factors affecting the success of crowdfunding were evaluated in this research regarding the initiators’ performance and the sponsors’ feedback, and the significant attributes were presented in the form of a decision tree structure using the data mining technique. the results reveal that the best performance of initiators is related to the field of direct investment attraction with 92% accuracy of the decision tree with the most important attributes of ”number of updates during the investment period” and ”number of dynamic technical and tactical analyses”.