Novice engineer vs. machine: comparing artificial neural network predictions with student estimates of market price using function structure models

Abstract

AbstractThis paper investigates the use of Artificial Neural Networks (ANN) to model human behaviour in early-stage engineering design; specifically, by comparing the predictive capability of ANNs and engineering novices. The comparison is between ‘crowds' of ANNs and humans. Function structure models of fifteen products are used as input for prediction. Additionally, the type of information provided is varied between topology information and vocabulary information. Prediction accuracy is compared, with the results showing that ANN crowd performs comparably to the novice estimates. However, students are more precise with their predictions. Moreover, student confidence in predictions is analyzed, and results suggest that students have low to moderate confidence in their predictions. Confidence increased with the addition of vocabulary information. Finally, limitations and future work are discussed, with research questions presented for subsequent research. This work motivates future studies on crowds, both human and virtual and the cost-benefits associated with collective intelligence.KEYWORDS: Wisdom of the crowdANN Crowdfunction structuredeep learningmarket value prediction Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 https://design.engr.oregonstate.edu/repo.