Neuro-Model for Improving the University-Industry Collaboration and Intellectual Property

Abstract

Some universities in the United States are leading technology transfer by their many close partnerships with government and industry. This has benefited them financially and by enhancing their research reputation. Patent-based intellectual property is a determinant factor, so an adequate cost-aware model must be derived to understand the process completely. This chapter presents the design and results of an artificial neural network (ANN) which relates the patent cost and the primary inputs of the process to model performance. Such inputs are invention disclosure, new patents issued, U. S. patents issued, licenses and optional executed, and other major agreements. A prediction of patent's cost could help a technology transfer office decide over the research to be patented but also to evaluate cost benefits. In addition, an integral solution is proposed where the positions of doctoral students and postdocs are defined. Overall, generating high quality invention disclosures is improved based on a more effective relationship between universities and industries.