To achieve the final objective of my Ph.D. work, which is to design hybrid models for disease prediction on health datasets (such as the Statlog or Cleveland Heart datasets), I need to analyze various data mining and machine learning algorithms. I am confused about which algorithms to choose and on what basis, and I plan to conduct a theoretical study before implementing and performing the analysis on tools such as WEKA or MATLAB. Any suggestions on this topic are welcome, and please provide any relevant links to related work if possible. Thank you.
Dear Aman,
At first, you have to consider that there is no panacea for all ML problems meaning that you have to select proper algorithms based on your dataset. For example, deep learning algorithms work well for categorical data and tree-based leaning algorithms work superior for tabular data. Moreover, if you have not-identically distributed unbalanced data you have to use federated learning including transfer learning. The story does not end here!
Sincerely,
Dear Aman Darolia
Check these articles. They will help you
Article Novel Feature Selection and Voting Classifier Algorithms for...
Article MbGWO-SFS: Modified Binary Grey Wolf Optimizer Based on Stoc...
For classification, check SVM or CNNS,
For segmentation check UNET(specifically for semantic segmentation)
- Badges
- Science topic