Machine Learning Platform for the Classification of Nucleotide Sequences
Université du Québec à Montréal


CASTOR-build
Use to create and train a classifier on your labeled sequences
For examples click on
 Load DNA sequences
 Abort upload

 {{uploadFileSeq.originalFileName}}  
 Remove the file 

 Load sequence labels
 Abort upload

 {{uploadFileAnnot.originalFileName}}  
 Remove this file 

Parameters 
Select metrics
[CUT] Number of fragments
[RMS] Root mean square of fragment lengths
Select features
By Spearman correlation. Coefficient threshold:
By Information gain. Top-K:                            
By a set of restriction enzymes
Training set options
Generate a negative dataset (beta version)
Size ratio of negative set and positive set
Select learning algorithms
Learn all available algorithms
[J48] Decision tree
[RFT] Random forest
[SVM] Support vector machine
[IBK] K-nearest neighbors (KNN)
[BAG] Bagging combined with J48
[ADA] AdaBoost combined with J48
[NBA] Naive Bayes
Classification metadata
Sequences taxon
Sequence type
Classifier description
Classification type
To run another job please refresh this page or click on Refresh

Please copy and keep the following link to access your result later:
http://castor.bioinfo.uqam.ca/index.php?castor=build&jobId={{theJobID}}