For the adequate treatment of patients, it is important to have an accurate and reliable algorithm developed for construction of a diagnosis system that can deal with gene expression data of DNA microarray, or proteomic data obtained by means of mass spectrometry (MS). It is also necessary that this algorithm is fast because these data consist of thousands of attributes (genes or proteins). We have developed a boosted fuzzy classifier with a SWEEP operator (BFCS) method on the basis of the fuzzy theory and boosting algorithm. This method has been applied for the construction of class predictors for cancer diagnosis using clinical data for breast cancer or proteomic pattern data of MS for ovarian cancer. The model performance has been evaluated by comparison with a conventional method such as a support vector machine (SVM) and a fuzzy neural network combined with the SWEEP operator (FNN-SWEEP) method previously proposed by us. The BFCS algorithm is 1,000 to 10,000 times faster than the other two methods. The constructed BFCS class predictors could discriminate classes of breast cancer and ovarian cancer with the same or higher accuracy than the other two methods. Furthermore, BFCS enabled the calculation of the reliability index for each patient, while the feature is not incorporated into a conventional algorithm. Based on this index, the discriminated group with 100% prediction accuracy was separated from the others.