Figure 5.

Comparison of three machine-learning approaches using various representations of the underlying features. Each filled circle is an R2 on a cross-validated test set, so there are 5 R2 values per method/feature combination.

Comparison of three machine-learning approaches using various representations of the underlying features. Each filled circle is an R2 on a cross-validated test set, so there are 5 R2 values per method/feature combination.