I am working on building a classification model for COVID19 from CT-scan series of COVID19 and non-COVID19 images, a relatively big dataset. What is your coding experience building similar models, mainly in terms of getting good f1-score? Do you recommend hybrid-model to get good results? have you experienced specific worthy techniques that have given you some interesting outcomes?
Dear Kenan Morani
,Look the link, maybe useful.
Article A multi-label classification model for full slice brain comp...
Regards,
Shafagat
Dear Kenan Morani
Have a look, these articles might be helpful:
1. https://towardsdatascience.com/chest-ct-scan-machine-learning-in-5-minutes-ae7613192fdc
2. https://www.nature.com/articles/s41598-021-87523-1
Kind Regards
Qamar Ul Islam
First of all, thank you for devoting your time to the research of Covid diagnosis in this time of need.
You have not clearly mentioned which dataset you are currently using. If you are using a publicly available dataset, then you might find some existing CNN models that have produced state-of-the-art results on that particular dataset. For that, you need to study some review paper. The following papers might help:
1) Article Systematic Review of Artificial Intelligence Techniques in t...
2) Article A review of mathematical modeling, artificial intelligence a...
3) https://ieeexplore.ieee.org/iel7/6287639/9312710/09351923.pdf
However, papers are coming up every day with new and novel methods. You should keep a track of them.
Also check https://link.springer.com/article/10.1007/s13755-021-00140-0
Hello look at this paper
https://www.nature.com/articles/s41598-021-87523-1
Dear Kenan Morani
Check this paper, it can help you
Article Novel Feature Selection and Voting Classifier Algorithms for...
this paper can help you
http://web.b.ebscohost.com/ehost/pdfviewer/pdfviewer?vid=2&sid=217c46a9-d915-4bd8-84a0-f6965c606bbf%40sessionmgr101
- Badges
- Science topic
- Similar topics
- Data Mining
- Classification