symbolic regression overfit problem!

[John-8911797110]

Thank your work again! 👍

I countered some questions in training, my envirenment is following:
My data is a dataframe like

     col0  col1  col2  col3  target
0     5.1   3.5   1.4   0.2       0
1     4.9   3.0   1.4   0.2       0
2     4.7   3.2   1.3   0.2       0
3     4.6   3.1   1.5   0.2       0
4     5.0   3.6   1.4   0.2       0
...   ...   ...   ...   ...      ...

I have 100+ cols and 1 target, dataframe shape is (200000, 100+), and I add 60 custom operators, But I found overfitting problem when setting large generations (10+ for me). Most individuals in population have similar express. So I wonder how to train models with overfitting control?

Appreciate your reply!! 😸