4 Replies Latest reply on Mar 3, 2017 1:37 PM by jgorsline RSS
    cxiao Newbie

    ThingworxAnalytics

    Hi Friends,

        I  have some problem about some parameters in this pictureA ,what these means  ? and what this paraments mean in pictureB and why should I fill these.

      • Re: ThingworxAnalytics
        cmorfin Creator

        Hi Joe

         

        For picture A, the table and the ROC curve goes together and described the accuracy of the model.

        You can maybe have a look at this blog for some input: Metrics for Model evaluation used in ThingWorx Analytics

         

        For picture B, you have got a good explanation if you select the online help in Builder on that page. (pick on the question mark icon)

        Basically if you select some important feature the scoring result will be appended with the number of important feature you selected.

        Those important features are the features that influence most the prediction of the goal.

        So this give you some added insight on what is impacting the prediction, but it does no change the prediction.

        The additional Features is used if you want to show the value of some specific features in the output scoring result. This again does not change the prediction, it just add some information inside the scoring result file.

         

        hope this helps

        Christophe