machine learning based rule learning and consistency checking from a complex decision table

[Dan]

Hello, 

I have a decision table with about 350 rows and 15 columns. There are various semantic (and functional) dependencies between values in different columns, and in particular the three last columns are outputs, given the values in the other columns as input. 

Values are sometimes binary, sometimes, one selection among a small list of options, sometimes, don't care, and always discrete. 

There are a lot of don't cares as well. One row, for example, has all but one value as don't care, leading to a specific output -- its a most generalized rule.

Are there any implemented rule learning algorithms in Prolog that could be used to analyze the decision table to come up with simpler symbolic rules, or even check for inconsistencies?

thanks,

Dan 

[Fabrizio Riguzzi]

Hello Dan,

please check Aleph and SLIPCOVER in cplint on swish

http://cplint.eu/example/aleph/aleph_examples.swinb

http://cplint.eu/example/learning/learning_examples.swinb

Both learn rules, the latter probabilistic ones

Best

Fabrizio

--
You received this message because you are subscribed to the Google Groups "SWI-Prolog" group.
To unsubscribe from this group and stop receiving emails from it, send an email to swi-prolog+...@googlegroups.com.
Visit this group at https://groups.google.com/group/swi-prolog.
For more options, visit https://groups.google.com/d/optout.

[Dan]

Hi Fabrizio, 

Thank you for the pointer -- somehow I had this feeling that a response will come from Italy :-)

Btw,  Aleph, (speaking  of mysticism) is also the first letter of the Hebrew word - Ein-sof  -- infinite (or, literally, no-end)...  quite, appropriately, perhaps, for an inductive reasoner   :-)

Dan

[Fabrizio Riguzzi]

Hi Dan,

Il giorno mer 23 gen 2019 alle ore 08:51 Dan <gros...@gmail.com> ha scritto:

Hi Fabrizio, 

Thank you for the pointer -- somehow I had this feeling that a response will come from Italy :-)

Are we famous for inductive reasoning? I didn't know that ;-) Thanks for the appreciation 

[Dan]

Hi Fabrizio,

I again and again notice very interesting work in logic in general and ontology and logic in particular originating from Italian research groups.

Dan