Relevance of Machine Learning
DOI link for Relevance of Machine Learning
Relevance of Machine Learning book
Today’s computer systems comprise a broad range of processors, communication networks, and information depositories. These systems are increasingly ubiquitous and, consequently, they are increasingly subject to attack, misuse, and abuse. The complexity of these systems makes it exceedingly difﬁcult to reason about their behavior. It is difﬁcult to design security policies that are simple to understand and ﬂexible to tolerate. Power and bandwidth limitations constrain security features in lightweight wireless devices. Cost considerations limit the usage of high-assurance implementation methods. Software-bundling policies make the software unwieldy; many vulnerable functions present in these bundles are rarely used by many users. System engineering tradeoffs are rarely based on
technology issues alone; social, organizational, economic, regulatory, and legal factors play a major role. Unfortunately, at today’s state-of-the-art, we do not have adequate understanding to develop an integrated solution to these challenging problems. The best we can do in the rest of the book is to present those technology issues we do understand and hope to lay the foundation so as to develop an ability to address the broader issues at a future time.