search

marine-debris / marine-debris.github.io

Quick Start Guide for MARIDA (Marine Debris Archive)
MIT License
41 stars 10 forks source link

Can't replicate random forest model #5

Closed EmanuelCastanho closed 1 year ago

EmanuelCastanho commented 1 year ago

I am trying to replicate the Random Forest model that uses Spectral Signatures and Spectral Indices, but I am unable to achieve the same recall (Pixel Accuracy) results.

I noticed that the results can change between Python versions depending on the packages, so I am using the same environment provided by your yml. My hyper-parameters are the same (including seed), I am also using the weights calculated from Confidence and Water Super Class. I am using the provided dataset.h5 and dataset_si.h5.

These are my results: results_hdf

As you can see, recall is slightly different from your PA results: 1: 0.92; 2: 0.93; 3: 0.92; 4: 0.27; 5: 0.7; 6: 0.82; 7: 0.83; 8: 1; 9: 0.48; 10: 0.83; 11: 0.33

Do you know what can be causing this, since the seeds and data are all the same?

Thank you!

Zhanarik commented 1 year ago

Hello everyone, I was trying to run random forest. This process took me one day, still can't achieve the dataset. python engineering_patches.py --type texture

Tried several times, it just keeps running without any information.

Wanted to download from the link, but the link seems broken. Could you please have a look?

Screen Shot 2023-02-16 at 16 33 21
EmanuelCastanho commented 1 year ago

Hi @Zhanarik , This step 2 takes a lot of time, but you have a progress bar that informs you. I have the texture folder saved on my computer, but it is quite heavy to send (3GB).

Please create a new issue about your problem, I will close this one because I solved my issue.

Zhanarik commented 1 year ago

Hi @EmanuelCastanho, I'm very glad that you found the answer for your question.

Thank you for notifying me.