Auto classification Error.Impossible to deal non boolean target variable .multiple categories found

[Former Member]

Hi Team

I am trying to classify data into multiple categories using Autoclassification in SAP PA 3.0 .

When I use only two categories the classification is done correctly but when the target variable categories increases the Expert analytics as well as automated analytics gives error.

I am not aware of any limitations on SAP PA to classify data into multiple categories. I believe I am doing somehting wrong . Please help me with this issue.

[AbdelhalimDadouche · Accepted solution]

Hi,

The Auto-Classification (used via Automated or Expert Analytics) is a "binary" classification and not a "multi-class" classification algorithm. Therefore, your target variable can only have 2 possible values.

If you want to address a "multi-class" classification scenario, you potentially have 2 strategies to choose from:

• One-vs.-rest (OvR)

The one-vs.-rest (or one-vs.-all) strategy involves training a single classifier per class, with the samples of that class as positive samples and all other samples as negatives. This strategy requires the base classifiers to produce a real-valued confidence score for its decision, rather than just a class label; discrete class labels alone can lead to ambiguities, where multiple classes are predicted for a single sample.

• One-vs.-one (OvO)

In the one-vs.-one, one trains K (K − 1) / 2 binary classifiers for a K-way multiclass problem; each receives the samples of a pair of classes from the original training set, and must learn to distinguish these two classes. At prediction time, a voting scheme is applied: all K (K − 1) / 2 classifiers are applied to an unseen sample and the class that got the highest number of “+1” predictions gets predicted by the combined classifier.

You can read the following blog post I wrote a while back:

https://blogs.sap.com/2016/03/09/how-to-handle-multiclass-classification-in-automated-analytics/

Regards

@bdel