Hi everyone, thanks for watching! Here's the link to the code on my Github in case you want to try it for yourself (also feel free to add me as a friend on Github if you want 🙂) : github.com/mikesaint-antoine/Comp_Bio_Tutorials/tree/main/melanoma_image_classifier
@carlovonterragon Жыл бұрын
Really good Project ! Implementing this on a website where you can just upload and classify your own skin-pictures ould be exciting.
@MikeSaintAntoine Жыл бұрын
Hey Carlo, yeah that's a cool idea! Although my guess is that there are probably a lot of legal and regulatory issues that would come up in terms of a website being used for actual medical advice. Also, the model would have to be way more accurate than the final version I came up with in this tutorial. I kept it simple for the purposes of teaching PyTorch, and got to about ~84% accuracy, but I think for use in an actual medical setting it would need to be like 99% accurate -- which I think could be achievable with some more clever tweaks to the model, combined with more training runs and possibly a larger dataset! My guess is that whoever actually WON the Kaggle competition for this dataset probably had a very accurate final model, potentially good enough for clinical use. Thanks for watching! 🙂
@JoDeNapole11 ай бұрын
Great project once again! Keep up the good work! I'll try changing the dataset to see how the same model structure will behave
@MikeSaintAntoine11 ай бұрын
Yeah I think that's a good idea. Btw there are a ton of good practice datasets to check out here: www.kaggle.com/datasets
@baggelis98 Жыл бұрын
Nice project! I will try to do the same with different cnn model for practice
@MikeSaintAntoine Жыл бұрын
Thanks! If you are trying to find a good dataset to practice with, there are a lot of great ones here: www.kaggle.com/datasets Let me know if you have any questions 🙂
@pedroTI2 ай бұрын
I want use your job ! This helping me in final academy work.
@MikeSaintAntoine2 ай бұрын
Let me know if you have any questions! 🙂
@pedroTI2 ай бұрын
@@MikeSaintAntoine Can you send me your email so I can contact you in a more professional manner?