Check the learning curves below, and confirm the state (high bias or high variance) of each algorithm.
The score of the learning curves represents F-measure (classification task) or negative mean squared error (regression task).
High score means the performance of the model (algorithm) is high.

If high performance models tend to be high variance, feature selection would help to get better results.
MALSS interactive supports feature selection.
If you want to try feature selection, press "Try feature selection".
If not, press "Continue".