CISIM 2023,Tokyu, 22-24 September 2023
Data Privacy, Energy Awareness and Credibility:
Advances in Machine Learning
Department of Electrical & Computer Engineering
University of Alberta, Edmonton, Canada
Over the recent years, we have been witnessing spectacular achievements of Machine Learning with highly visible accomplishments encountered, in particular, in natural language processing and computer vision impacting numerous areas of human endeavours. Driven inherently by the technologically advanced learning and architectural developments, Machine Learning constructs are highly impactful coming with far reaching consequences; just to mention autonomous vehicles, control, health care imaging, decision-making in critical areas, among others.
We advocate that the design and analysis of ML constructs have to be carried out in a holistic manner by identifying and addressing a series of central and unavoidable quests coming from industrial environments and implied by a plethora of requirements of interpretability, energy awareness (being also lucidly identified on the agenda of green AI), efficient quantification of quality of ML constructs, their brittleness and conceptual stability coming hand in hand with the varying levels of abstraction. They are highly intertwined and exhibit relationships with the technological end of ML. As such, they deserve prudent attention, in particular when a multicriterial facet of the problem is considered.
The talk elaborates on the above challenges, offers definitions and identifies the linkages among them. In the pursuit of coping with such quests, we advocate that Granular Computing can play a pivotal role offering a conceptual environment and realizing algorithmic development. We stress and identify ways to effective assessments of credibility of ML constructs. As a detailed study, we discuss the ideas of knowledge transfer showing how a thoughtful and prudently arranged knowledge reuse to support energy-aware ML computing. We discuss passive and active modes of knowledge transfer. In both modes, the essential role of information granularity is identified. In the passive approach, information granularity serves as a vehicle to quantify the credibility of the transferred knowledge. In the active approach, a new model is constructed in the target domain whereas the design is guided by the loss function, which involves granular regularization produced by the granular model transferred from the source domain. A generalized scenario of multi-source domains is discussed. Knowledge distillation leading to model compression is also studied in the context of transfer learning.