It has been known that either linear correlation or nonlinear correlation might exist between featureto-feature and feature-to-class in datasets. In this paper，we study the differences of selected feature subset whendifferent kinds of measures are applied with same feature selection method in different kinds of datasets. Threerepresentative linear or nonlinear measures，linear correlation coefficient，symmetrical uncertainty，and mutualinformation are selected. By combining them with the fast correlation-based filter ( FCBF) feature selectionmethod，we make the comparison of selected feature subset from 8 gene microarray and image datasets.Experimental results indicate that the feature subsets selected by linear correlation coefficient based FCBF obtainbetter classification accuracy in gene microarray datasets than in image datasets，while mutual information andsymmetrical uncertainty based FCBF tend to obtain better results in image datasets. Moreover，symmetricaluncertainty based FCBF is more robust in all datasets.