Author: Waymond Rodgers

Dominant Algorithms to Evaluate Artificial Intelligence: From the View of Throughput Model

eBook: US $79 Special Offer (PDF + Printed Copy): US $145
Printed Copy: US $105
Library License: US $316
ISBN: 978-981-5049-55-8 (Print)
ISBN: 978-981-5049-54-1 (Online)
Year of Publication: 2022
DOI: 10.2174/97898150495411220101

Introduction

This book describes the Throughput Model methodology that can enable individuals and organizations to better identify, understand, and use algorithms to solve daily problems. The Throughput Model is a progressive model intended to advance the artificial intelligence (AI) field since it represents symbol manipulation in six algorithmic pathways that are theorized to mimic the essential pillars of human cognition, namely, perception, information, judgment, and decision choice. The six AI algorithmic pathways are (1) Expedient Algorithmic Pathway, (2) Ruling Algorithmic Guide Pathway, (3) Analytical Algorithmic Pathway, (4) Revisionist Algorithmic Pathway, (5) Value Driven Algorithmic Pathway, and (6) Global Perspective Algorithmic Pathway.

As AI is increasingly employed for applications where decisions require explanations, the Throughput Model offers business professionals the means to look under the hood of AI and comprehend how those decisions are attained by organizations.

Key features:

  • - Covers general concepts of Artificial intelligence and machine learning
  • - Explains the importance of dominant AI algorithms for business and AI research
  • - Provides information about 6 unique algorithmic pathways in the Throughput Model
  • - Provides information to create a roadmap towards building architectures that combine the strengths of the symbolic approaches for analyzing big data
  • - Explains how to understand the functions of an AI algorithm to solve problems and make good decisions
  • - Informs managers who are interested in employing ethical and trustworthiness features in systems.

Dominant Algorithms to Evaluate Artificial Intelligence: From the view of Throughput Model is an informative reference for all professionals and scholars who are working on AI projects to solve a range of business and technical problems.

Audience: Working professionals, data scientists, engineers and project managers interested in AI models for business projects; computer scientists and Computer science students.

Preface

“Artificial intelligence would be the ultimate version of Google. The ultimate search engine that would understand everything on the web. It would understand exactly what you wanted, and it would give you the right thing. We're nowhere near doing that now. However, we can get incrementally closer to that, and that is basically what we work on.” —Larry Page

I think we're going to need artificial assistance to make the breakthroughs that society wants. Climate, economics, disease -- they're just tremendously complicated interacting systems. It's just hard for humans to analyze all that data and make sense of it.

Artificial intelligence (AI) systems are already transforming individuals and organizations manner of functioning in today’s world. AI can automate repetitive tasks, analyze large volumes of data, recommend content, translate languages, and even play games. Further, AI and related technologies are progressively ubiquitous in business and society. For example, AI increasingly find their way into everything from advanced quantum computing systems, automobiles, household appliances, and leading-edge medical diagnostic systems to consumer electronics and “smart” personal assistants. AI tools are also employed virtual reality, augmented reality as well as making IoT devices and services smarter and more secure.

Nonetheless, the current scope of things that AI can accomplish is relatively narrow. Some experts say the technology is far from becoming so-called artificial general intelligence, or AGI. That is, AGI is the capability to understand or learn any intellectual task that a human being can.

Furthermore, others have noted that even in its current, narrow proficiencies, AI provokes a series of ethical and trustworthiness questions. These questions represent issues such as whether the data fed into AI programs are without bias, and whether AI can be held accountable if something goes wrong.

To construct ethical and trusted AI systems, there needs to be cooperation among nations and various stakeholders. Experts have previously warned that inherently biased AI programs can present momentous problems and it may get in the way people’s trust in those systems. For example, facial recognition software, for example, may incorporate accidental racial and gender bias, which may pose a threat to a particular group of individuals.

Therefore, this book provides a methodology described as the Throughput Model that can enable individuals and organizations to better identify, understand, and use algorithms to solve daily problems. Moreover, the Throughput Model can further the AI field since it represents symbol manipulation in six algorithmic pathways that seems to be essential for human cognition, namely, perception, information, judgment, and decision choice. Finally, The Throughput Model provides the first steps towards building architectures that combine the strengths of the symbolic approaches that can be adapted for machine learning/deep learning, and to develop better techniques for extracting and generalizing abstract knowledge from large, often noisy data sets.

As AI is employed more and more for applications where decisions require explanations, the Throughput Model offers the means to look under the hood of AI and comprehend how those decisions are attained by organizations. This is, particularly important for employing ethical and trustworthiness systems. Hence, Throughput Modelling ought to be considered from the start as it will inform the design of an AI system. Building trusted and ethical AI systems and the governance around them may potentially become a competitive strength for organizations.

CONSENT FOR PUBLICATION

Not applicable.

CONFLICT OF INTEREST

The authors declare no conflict of interest, financial or otherwise.

Waymond Rodgers
University of Hull
University of Texas
El Paso
USA