I have modelled a simple regression tree on KNIME. When I use a 70-30 percent train test data, I get an acceptable R Squared of 0.9. but when I use X-Partitioner and X-aggregator for cross-validation the R-squared becomes 1. What is the problem that I get 1?