چكيده به لاتين
Eye movement studies are widely used in human cognition. Brain activity and memory load impress eye movement. Here, we investigate whether fluid intelligence (FI) has any effect on eye movement pattern in a given task. Using Cattell test, FI of individuals were measured and divided into three groups: low FI, middle FI, and high FI. Eye movement of individuals were recorded while they were trying to find the number of differences between two almost identical images. Fixation duration, number of fixation, fixation distance, and saccade average velocity of three groups were extracted and compared statistically. Our experiment demonstrated that fixation duration of low FI individuals was greater than middle FI and high FI individuals. In contrast, the number of fixation, fixation distance and saccade average velocity was larger for subjects with higher FI. To further study the effects of subjects’ FI on eye movement, we have performed our experiment based on 2 scenarios. In the first scenario, we extracted 10 features and categorized individuals into three group by support vector machine (SVM) classifier. SVM obtained the value of 63.4% accuracy, 58% sensitivity, 81.7% specificity, and 59.2% F1-Score. In the second scenario, we extracted 24 features and computed Pearson correlation coefficient (PCC) between each features and FI in order to select the best features. Then, we predicted individuals’ FI by support vector regression (SVR) with 0.69 PCC between observe FI and predicted FI. In our view the result emphasizes influences of FI on eye movement during visual search.