R2 as the metrics

[ZeqingBao]

evaluation-metrics/evaluation_metrics.py

Line 185 in 68a2be8

r2 = r2_score(

Maybe we can consider using multiple metrics instead of relying on just one, as a single metric can introduce potential bias. We could look at commonly used ones like RMSE, MAE, and PCC to get a more well-rounded evaluation.

[ZeqingBao]

I assume this is for single-objective optimization (apologies if I missed that).

In any case, it is probably also worth discussing how we evaluate model accuracy across multiple objectives.