EXPLAINABLE MACHINE LEARNING FOR ETHICAL ARTIFICIAL INTELLIGENCE BASED DECISIONS

Radu Stefan () and George Carutasu ()
Additional contact information
Radu Stefan: Politehnica University of Timisoara, Timisoara, Romania
George Carutasu: Romanian-American University, Bucharest, Romania

Journal of Information Systems & Operations Management, 2020, vol. 14, issue 1, 151-161

Abstract: In the last century, many approaches and tools have been developed to implement systems that would achieve Artificial Intelligence (AI) and solve most difficult problems in computer science with that. The quest is to find the most suitable machine learning method for a given problem domain. Currently the approach that yields the best results in practice in most domains is based upon artificial neural networks, or the more advanced deep neural networks. Neural networks are highly efficient for most common scenarios, e.g. language understanding, image recognition and the like. Typically, models are trained for a specific task, and their performance is judged based on the binary outcome, e.g. in image recognition, whether the artefact was recognized or not. From an ethical decision-making point of view, the challenge remains to identify how the learning process influences the outcome and finally to rationally understand how a decision has been made. In this paper we present a framework for an enhanced machine learning model, that adds a layer of decision explanation to the outcome towards the user of the system. We suggest how the learning model needs to be expanded to an explainable model and the corresponding explainable interface that would allow presentation towards the user.

Date: 2020
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.rebe.rau.ro/RePEc/rau/jisomg/SU20/JISOM14.12020_151-161.pdf (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:rau:jisomg:v:14:y:2020:i:1:p:151-161

Access Statistics for this article

More articles in Journal of Information Systems & Operations Management from Romanian-American University Contact information at EDIRC.
Bibliographic data for series maintained by Alex Tabusca ().