Support Vector Machine Classification of Microarray Data

摘要: The Problem: Use the learning from examples paradigm to make class predictions and infer genes involved in these DNA microarray expression data. Specifically, we use a Support Vector Machine (SVM) classifier [6] predict cancer morphologies treatment success determine relevant inference. Motivation: Previous Work: A generic approach classifying two types of acute leukemias was introduced Golub et. al. [3]. SVM’s have been applied this problem [5] also predicting functional roles uncharacterized yeast ORF’s [1]. Approach: We used SVM discriminate between leukemia. output classical is designation ±1. In particular application it important be able reject points for which not confident enough. confidence interval on that allows us with low values. It are classification. preliminary results feature selection algorithm classifiers. trained 38 training set tested 34 test set. Our (see table 2 figure (1)) best reported so far dataset.