FINAL Rules of Evolutionary Forest about evolutionaryforest

[hengzhe-zhang]

Hi! Thanks for your interest in this package. It is possible to visualize the final result by printing all GP individuals in the hall-of-fame. However, a better way might be to only visualize important features because Evolutionary Forest may contain hundreds of high-order features. To visualize important features, please refer to a tutorial.

[MohammadSoltani100]

Hello dear Hengzhe Zhang Thank you so much for the strong package and your kind reply. In fact, my goal is to model a hydrological process, in which both 'important features', and final 'if-then rule' are important. For your information, I want to use both "PSTree" and "Evolutionary Forest" for my purpose. my question is how can I reveal the final expression (script)? Thank you very much for your valuable work. best regards

…

On Sun, May 22, 2022 at 3:56 PM Hengzhe Zhang ***@***.***> wrote: Hi! Thanks for your interest in this package. It is possible to visualize the final result by printing all GP individuals in the hall-of-fame. However, a better way might be to only visualize important features because Evolutionary Forest may contain hundreds of high-order features. To visualize important features, please refer to a tutorial. <https://github.com/hengzhe-zhang/EvolutionaryForest/blob/master/tutorial/diabetes_regression.ipynb> — Reply to this email directly, view it on GitHub <#74 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ASPWCXB3WUPDEGXSTEWLELTVLIKYDANCNFSM5WR6B2XA> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[hengzhe-zhang]

Hi @MohammadSoltani100 ! Here is a way to print the final expression. However, the final decision tree might be large and unexplainable. A more advanced approach for interpretable-ML is still under-developed.

for id, h in enumerate(regr.hof):
    print('Tree', id)
    print('GP Features', h)
    print(tree.export_text(h.pipe['Ridge']))

[MohammadSoltani100]

You are right Wait for that Thanks a lot.

…

On Tue, May 24, 2022, 7:56 AM Hengzhe Zhang ***@***.***> wrote: Hi @MohammadSoltani100 <https://github.com/MohammadSoltani100> ! Here is a way to print the final expression. However, the final decision tree might be large and unexplainable. A more advanced approach for interpretable-ML is still under-developed. for id, h in enumerate(regr.hof): print('Tree', id) print('GP Features', h) print(tree.export_text(h.pipe['Ridge'])) — Reply to this email directly, view it on GitHub <#74 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ASPWCXF2E7UHP23RLRFTLWDVLRD4VANCNFSM5WR6B2XA> . You are receiving this because you were mentioned.Message ID: ***@***.***>