My name is Douglas Rofes, Iím a student at CUI, (university of Geneva) I used a simple algorithm: NAIVE BAYE, I added some modification with the main object of better to adapt it to data given. For instance I use a combination of Majory Voting and Bayes Formulas, for the selection of the best probable values in the histograms of distribution. Herewith, some results for dataset V s1 | D1 / D2 = .6532 S1 | D2 / D3 = .7209 s1 | D1 + D2/ D3 = .7054 s2 | D1 / D2 = .5451 S2 | D2 / D3 = .6296 s2 | D1 + D2/ D3 = .6393 s3 | D1 / D2 = .4200 S3 | D2 / D3 = .4026 s3 | D1 + D2/ D3 = .4297