I would like to implement the support vector machine-recursive feature elimination algorithm to a biomedical dataset. I am a beginner in machine learning, and would like to know how I can implement this algorithm.
I currently have a proteomics dataset comparing two cell lines (WM and 1205) that have been analyzed in triplicate. I was hoping to apply the support vector machine with recursive feature elimination to classify on the basis of cell line. Is this algorithm appropriate for data of this type, or are more samples (I only have 3 per cell line) necessary for accurate classification? What would be the appropriate method to train the SVM? I was thinking that I would determine the distribution of the data, and generate a simulated dataset on which to train the SVM. Does the training set need labels (e.g. {1, -1})?
Any help would be greatly appreciated.
