Industry-University collaborations are known to offer firms a competitive edge by giving them access to a diverse range of complimentary resources outside the confines of their organisations. However, failure in these collaborations is a common problem. Failure can result in a waste of resources delays in meeting customer demands and therefore it is undesirable. One major risk category associated with failure is psychological risks. The concept of change triggers negative responses amongst employees and therefore can affect the performance of these collaborations. Detecting these risks early can help managers mitigate them and take steps necessary to ensure success. In this paper, an Artificial Neural Network approach is used to reduce industry-university collaboration failure by predicting failure. The Artificial Neural Network employed to build the models is the Multi-perceptron neural network and it is applied to two different datasets, one imbalanced and the other one balanced through means of a the Synthetic minority oversampling technique algorithm. The results reveal that the Multi-perceptron model built with a balanced dataset are better at predicting industry-university collaboration failure as compared to the one built with the raw imbalanced dataset.
Uzapi Hange (2022). An Artificial Neural Network Model for Industry-University Collaboration Failure Prediction Based on Psychological Risk Factors. Multidisciplinary International Journal of Research and Development (MIJRD), Volume: 02 Issue: 01, Pages: 61-70. https://www.mijrd.com/papers/v2/i1/MIJRDV2I10006.pdf