Figure 1From: Determining relative importance of variables in developing and validating predictive models Relative importance of variables for the prostate cancer data, measured using the percentage of selection out of 1000 bootstrap samples.Back to article page