Build and evaluate class predictors (classifiers) using the
K-nearest-neighbors (KNN) class prediction method.
What would you like to do?
Build and evaluate a classifier
Build and evaluate class predictors (classifiers) using the
K-nearest-neighbors (KNN) class prediction method. How many gene expression data sets do you have?
- Two data sets. Click here to
build a classifier against one data set and evaluate it against the other.
- One large data set. If you have one large data set, use the
SplitDatasetTrainTest
module to split it into two data sets. Then, click here to
build a classifier against one data set and evaluate it against the other.
- One data set. If you have one smaller data set that cannot easily be split, click here to
build and evaluate classifiers using cross-validation.
Evaluate an existing classifier
To evaluate an existing KNN classifier, you must have built the classifier using the KNN module. Do you have a classifier (*.knn.model) file?
Classify an unknown sample using an existing classifier
To classify an unknown sample, you must have built a KNN classifier using the KNN module. Do you have a classifier (*.knn.model) file?