[Event at CIG] ADEMAL worshop

Miki Hermann hermann at lix.polytechnique.fr
Tue Mar 14 13:30:26 CET 2023


				ADEMAL
	       Automated Deduction for Machine Learning
	     https://sites.google.com/view/ademalcade2023

		  Workshop affiliated with CADE 2023
		 https://easyconferences.eu/cade2023/

		   Rome, Italy          5 July 2023



Machine Learning on Big Data  has gained an enormous popularity during
recent  years. In  the vast  majority of  cases, the  machine learning
software  for  a  particular  data  set is  constructed  as  a  neural
network. However, this approach is  in need of methods for explainable
artificial intelligence, as most of  the applied models are opaque and
operate as  a black-box.  Other possible  issues are  the reliability,
robustness and fairness of the machine learning software. These issues
could be addressed with formal methods from automated deduction.

The purpose of this workshop is to bring together researchers who work
in  machine  learning and  automated  deduction.  The targets  of  the
workshop  will be:  to apply  automated deduction  methods to  machine
learning, to produce objects recognizable by automated deduction tools
by  machine  learning  methods,   to  enhance  explainable  artificial
intelligence  coupled with  machine  learning  by automated  deduction
tools, and  to consider  further developments  in the  intersection of
both research areas.


Nonexclusive list of topics:
 - machine learning in general
 - automated deduction methods applied to machine learning
 - theoretical foundations of explainable AI
 - verifying explanations
 - verifying deep architectures
 - metrics for robustness and explanation quality
 - neurosymbolic reasoning
 - logic, calculi, and algebras in explainable AI
 - machine learning systems using automated deduction methods


We seek four types of submissions:
 - full research papers of up to 15 pages + 3 pages of references;
 - short communications of up to 5 pages;
 - tool descriptions of up to 15 pages + 3 pages of references;
 - and recent previously published work (no publication, presentation
only).

The  tool descriptions  will  have to  contain a  link  to a  publicly
accessible server, such as github or similar, from where the described
software will be made available. The accepted papers will be published
in the OASIcs series (very probably as a post-conference proceedings).

Submissions will be handled by EasyChair conference management system.
https://easychair.org/conferences/?conf=ademal2023

Submissions  must be  formated  according to  the OASIcs  instructions
available at https://submission.dagstuhl.de/series/details/5#author .


Important dates:
 o Paper submission:  8 May 2023
 o Notification:     26 May 2023
 o Final version:    31 May 2023 (firm)


Participation:

One of the authors of each  accepted paper or tool description will be
expected  to present  it during  a 30  minutes slot,  out of  which 25
minutes will  be allocated to  the presentation  and 5 minutes  to the
subsequent discussion.


Program Committee

   Martin C. Cooper    (IRIT, Toulouse, France)
   Miki Hermann        (LIX, Ecole Polytechnique, Palaiseau, France)
   Francesco Leofante  (Imperial College, London, UK)
   Joao Marques-Silva  (IRIT, Toulouse, France)
   Nina Narodytska     (VMware Research, Palo Alto, California, USA)
   Corina S. Pasareanu (NASA Ames Research Center, CyLab,
                        Carnegie Mellon University,
			Mountain View, California, USA)
   Aditya A. Shrotri   (Rice University, Houston, Texas, USA)



More information about the IFI-CI-Event mailing list