Question

Beginners bioinformatics project with machine learning?

0

Entering edit mode

23 months ago

godwin ▴ 30

I need ideas or pointers to build a bioinformatics project using machine learning

python machine-learning • 1.8k views

ADD COMMENT • link 23 months ago by godwin ▴ 30

score 0 · Answer 1 · 2022-05-05

0

Entering edit mode

23 months ago

Jeremy ▴ 880

One idea would be to predict clinical outcome using a Kaggle dataset. For example, see the following dataset: Lung Cancer

ADD COMMENT • link 23 months ago by Jeremy ▴ 880

1

Entering edit mode

Thank you..I'll be sure to check it out

ADD REPLY • link 23 months ago by godwin ▴ 30

0

Entering edit mode

I did this today with the dataset. Can you go through it and see if it's okay?

https://github.com/Godwin-Ani/Biologic/tree/main/Lung%20Cancer

ADD REPLY • link 23 months ago by godwin ▴ 30

0

Entering edit mode

This looks good to me. I use R for machine learning, although I believe Python is more popular. I would check with the people on Kaggle to see what they think.

ADD REPLY • link 23 months ago by Jeremy ▴ 880

1

Entering edit mode

Thank you... I'll be waiting for your feedback

ADD REPLY • link 23 months ago by godwin ▴ 30

0

Entering edit mode

You can try to optimize your model by adjusting the parameters. See what happens if you set n_estimators to 500 or 1000. You can also adjust max_features. The default is 'sqrt', which should be 4 in your case. You can try setting it to 1, 'log2', 2, and 8, as well. Definitely try max_features = 1. Again, you can post a similar question on Kaggle to see what they say.