Comparing RLS and SVM Classifiers for Cancer Classification
Author Information
Author(s): Nicola Ancona, Rosalia Maglietta, Annarita D'Addabbo, Sabino Liuni, Graziano Pesole
Primary Institution: Istituto di Studi sui Sistemi Intelligenti per I'Automazione, CNR
Hypothesis
Can Regularized Least Squares classifiers perform comparably to Support Vector Machines in cancer classification using DNA microarray data?
Conclusion
RLS classifiers are a valuable alternative to SVM classifiers for cancer classification due to their simplicity and low computational complexity.
Supporting Evidence
- RLS classifiers showed comparable performance to SVM classifiers in cancer classification.
- RLS classifiers require less computational resources than SVM classifiers.
- Leave-One-Out error was used to assess the generalization ability of both classifiers.
Takeaway
This study shows that a simpler method called RLS can classify cancer types just as well as a more complex method called SVM, making it easier to use.
Methodology
The study compared the performance of RLS and SVM classifiers using Leave-One-Out error on three different cancer datasets.
Limitations
The study focused only on two-class classification problems and did not explore multi-class scenarios.
Participant Demographics
The study involved cancer tissue samples from patients with different types of cancer.
Statistical Information
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website