mod unl lab exa ori -what do these mean?

onyourmark

Hi Guys. I have looked on the internet, in the documentation and in the forums. I am sure there must be an explanation in there somewhere but I could not find it. Can someone tell me what these mean and/or point me to some documentation that explains these and how they should be used? These are labels on the inputs and outputs on the operators.

Thanks!

cbwq

They're just short hand for different set types.

mod = Model
unl = Unlabelled data
lab = Labelled data
exa = Example set
ori = Original data

There's also another one I've seen "pre" which is a preprocessing model output by normalisation operators.

A Model is a set of weights or something similar, usually a vector and is typically output by learners of various types and required by the Apply Model operator.
Labelled data is the example set marked up with an additional attribute (column) with the role of label.  Unlabelled data is data without a label attribute.

Example data is the dataset that you're processing, usually.

And Original is a copy of the unmodified data that some operator is making changes to.  Usually ORI appears after an EXA, such as with the aggregate operator.  You can use it to make sure you're doing what you intended.

The wiki entries on these will list the inputs and outputs (it should be visible in the design perspective) and you can work out what's going on and what's required.  Use the Apply Model operator as an example, it has most of these different types.

cbwq.

onyourmark

Oh. Excellent. I will read what you have written carefully. Thanks.