Class Prediction

Click the desired algorithm

• CART (Breiman et al., 1984) builds Classification And Regression Trees. It works by recursively splitting the feature space into a set of non-overlapping regions and then predicting the most likely value of the dependent variable within each region. A classification (or regression) tree represents the set of nested if-then conditions used to predict a categorical dependent (or continuous dependent) variable based on the observed values of the feature variables. CART is vulnerable to overfitting and therefore not commonly used with microarray data.
• K-nearest-neighbors (KNN) classifies an unknown sample by assigning it the phenotype label most frequently represented among the k nearest known samples (Golub and Slonim et al., 1999). In GenePattern, an analyst can select a weighting factor for the 'votes' of the nearest neighbors. For example, one might weight the votes by the reciprocal of the distance between neighbors.
• Probabilistic Neural Network (PNN) calculates the probability that an unknown sample belongs to a given set of known phenotype classes (Lu et al., 2005; Specht, 1990). The contribution of each known sample to the phenotype class of the unknown sample follows a Gaussian distribution. PNN can be considered as a Gaussian-weighted KNN classifier - known samples close to the unknown sample have a greater influence on the predicted class of the unknown sample.
  
  PNN is not on the GenePattern public server. The PNN modules require the Windows operating system. To use PNN, install the GenePattern server and the PNN modules on a Windows machine.
• Support Vector Machines (SVM) is designed for multiple class classification (Rifkin et al., 2003). The algorithm creates a binary SVM classifier for each class by computing a maximal margin hyperplane that separates the given class from all other classes; that is, the hyperplane with maximal distance to the nearest data point. The binary classifiers are then combined into a multiclass classfier. For an unknown sample, the assigned class is the one with the largest margin.
• Weighted Voting (Slonim et al., 2000) classifies an unknown sample using a simple weighted voting scheme. Each gene in the classifier 'votes' for the phenotype class of the unknown sample. A gene's vote is weighted by how closely its expression correlates with the differentiation between phenotype classes in the training data set.