it’s my first time making this type of projects, i would love to get you guys opinions and feedback
(Github user name= H1potenus )Lung_Cancer_Detection that’s the program i’ve mentioned. i also have another project in there about apples. please tell me if there is anything i can improve,
I also tried to scale the data just for the age but i couldnt do it, so i had to make it for the whole data H1potenus (H1potenus) / Repositories · GitHub
Hello @H1potenus
it would be more approachable if you could share your GitHub link here for other to access your work.
Regards
DP
1 Like
this website didnt allow me to share links but my github nickname is H1potenus
(H1potenus (H1potenus) / Repositories · GitHub), oh i see, i can share it now, intresting
1 Like
Hello @H1potenus
Discourse Community allows you to share your GIthub link, only sharing solution assignment which are graded is not allowed in your Github repository as it is against community guidelines.
Regards
DP
1 Like
thank you for informing me, i m really new to this. i’ve updated my post
2 Likes
Will check your Github later and let you know, but my response might be delayed. Sorry.
1 Like
Hello @H1potenus,
Seems your dataset has a set of pretty imbalanced labels. I suggest you to google with keywords like “class imbalance metrics” and study about this problem, then perhaps you will come up with a different metric that better reflects the performance of your model.
Cheers,
Raymond
4 Likes
As @rmwkwok suggested, the dataset’s labels are overwhelmingly 1, which is typical when dealing with diagnostic data. Simple accuracy for a dumb trainer can be ~99% just by commanding it to predict 1 no matter what. You need more nuanced metrics. See below.
4 Likes
thank you and @Michael_Lee3 for your answers. i will look into it
2 Likes