Using machine learning models to classify user performance in the ruff figural fluency test from eye-tracking features

摘要

The Ruff Figurai Fluency Test is a paper and pencil tool to gain information about the nonverbal capacity for such activities as initiation, planning, and divergent reasoning including strategy use. It is applied voluntarily in the form of cognitive test batteries. In this study a computerised version of the Ruff Figural Fluency Test was employed in order to assess user cognitive performance. Sixty-one male participants were examined using the eye-tracking technique to gain the desired data. Different machine learning models were applied in order to classify user performance. The best results (78,7% for the testing dataset) were obtained for Quadratic Discriminant Analysis classifier.