Weka java classification class or not -
already done: in code there 10 flowers , have various data on 1 flower. trained data weka , when test it, gives correct answer new data ( if new test data applied model, give flower name)
classifiers.add(new naivebayes()); classifiers.add(new naivebayesupdateable()); (classifier c : classifiers) { double starttime = system.currenttimemillis(); classifier cls = c; cls.buildclassifier(train); evaluation eval = new evaluation(train); eval.evaluatemodel(cls, test); double estimatedtime = system.currenttimemillis() - starttime; txt.settext(txt.gettext() + classutils.getclassname(cls.getclass()) + " c: " + eval.correct() + " i: " + eval.incorrect() + " cp: " + 100 * (eval.correct() / (eval.correct() + eval.incorrect())) + "% t:"+estimatedtime/1000+"\n"); } want do: don't want know flower among 10 flowers. if input new data set, want know specific flower or not. can above method , if flower match , give true. take more processing power. want minimize processor usage.
wiki
Comments
Post a Comment