search

Alex-Lekov / AutoML_Alex

State-of-the art Automated Machine Learning python library for Tabular Data
MIT License
225 stars 42 forks source link

does it work on spectral data? #17

Closed fcoppey closed 3 years ago

fcoppey commented 4 years ago

Hello,

I'm wondering if you have tested it on spectral data? I have NIR Spectroscopy data with 125 variables and using presently TPOT, works very well but looking at your benchmark it seems your approach works better for the kind of data you have developed it for. hence do you recommend testing it on my spectral data or it's not the purpose of it?

thanks

Alex-Lekov commented 4 years ago

please send an example of data

fcoppey commented 4 years ago

Here you,

92 spectra of about 25 samples (several replicates spectra per sample). Two sheets, first is raw spectra, second is pre-treated, I usually give pre-treated spectra to TPOT.

Y column is target value for regression, then there 125 variables that represent the near infrared spectra with corresponding wavelengths.

Let me know if you need anything else.

Thanks

Florentin

On 30 Sep 2020, at 17:20, Alex Lekov notifications@github.com<mailto:notifications@github.com> wrote:

please send an example of data

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/Alex-Lekov/AutoML_Alex/issues/17#issuecomment-701460169, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AKFMG6LBRCU4XF4ULUCOSKLSINEEZANCNFSM4RUHYOEQ.

Alex-Lekov commented 3 years ago

deadline expired

fcoppey commented 3 years ago

Ok so you didn't have time to try? I sent you the data it seems.