Artificial Intelligence in a Throughput Model

Some Major Algorithms

Waymond Rodgers

• Publisher: CRC Press
• Year: 2020
• ISBN:
  9780429561030
• Edition: 1
• Topic: Computer Science
• Subtopic: Artificial Intelligence (AI) & Semantics
• Index: Computer Science

Physical and behavioral biometric technologies such as fingerprinting, facial recognition, voice identification, etc. have enhanced the level of security substantially in recent years. Governments and corporates have employed these technologies to achieve better customer satisfaction. However, biometrics faces major challenges in reducing criminal, terrorist activities and electronic frauds, especially in choosing appropriate decision-making algorithms. To face this challenge, new developments have been made, that amalgamate biometrics with artificial intelligence (AI) in decision-making modeling. Advanced software algorithms of AI, processing information offered by biometric technology, achieve better results. This has led to growth in the biometrics technology industry, and is set to increase the security and internal control operations manifold.

This book provides an overview of the existing biometric technologies, decision-making algorithms and the growth opportunity in biometrics. The book proposes a throughput model, which draws on computer science, economics and psychology to model perceptual, informational sources, judgmental processes and decision choice algorithms. It reviews how biometrics might be applied to reduce risks to individuals and organizations, especially when dealing with digital-based media.

1

Introduction to Artificial Intelligence and Biometrics Applications

Artificial Intelligence is the science and engineering of making intelligent machines, especially intelligent computer programs.

—John McCarthy, Father of Artificial intelligence

“Man is only great when he acts from passion.”

—Benjamin Disraeli (former Prime Minister of Britain)

Artificial intelligence (AI) is here for stay. For individuals and organizations, Artificial Intelligence is a disruptor in the way we live, learn, work and adopt to different situations. Disruptors represent a person, place or thing that prevents something, especially a system, process, or event, from enduring as customary or as anticipated in the future. We are right in the midst of the information revolution. Although it is an extraordinary time and place to be in, there are caveats that come along with it. Having a machine tell you how long your commute will be, what music you should listen to, and what content you would likely engage with are all relatively innocuous examples.

Artificial Intelligence is presently used as a tool to assist people. It is adopted as point solutions across a wide array of functions such as personal digital assistant, email filtering, search, fraud prevention, engineering, marketing models, digital distribution, video production, news generation, play and game-play analytics, customer service, financial reporting, marketing optimization, energy cost management, pricing, inventory, enterprise applications, etc. Artificial Intelligence is also integrated into biometrics tools such as iris recognition, voice recognition, facial recognition, content classification, gait, and natural language.

In addition, Artificial Intelligence is becoming widespread in most facets of decision-making and will become more so in the near future. Artificial Intelligence has been an aspiration of computer scientists since the 1950s, and has experienced colossal advancement in recent years. Artificial Intelligence implementation is already an integral part of many of our online activities and will become progressively more entrenched in everything we touch, hear, see and do. On a task-by-task basis, Artificial Intelligence systems gradually produce outputs that far exceed the precision and reliability of those produced by individuals. For example, pharmaceuticals and the food industry utilize Artificial Intelligence sensor tools to ensure the optimum temperature for creating drugs or cooking food. Other sensors make certain that products are stored and shipped at safe temperatures.

In agriculture, the implementation of Artificial Intelligence tools has boosted crop yields. Artificial Intelligence has provided for farmers to make better decisions pertaining to crops to sow and how to best manage them. Harvesting everything from grains to root vegetables to soft fruits can now be performed more efficiently and effectively with robots than people. Furthermore, mobile technology has reformed the manner in which field service teams operate. They can get work orders more rapidly, and once the employees are at a particular location, they can access schematics and documents to assist with the repair work. In addition, the Artificial Intelligence system provides for a smoother progression of work orders from fault call to the ordering of parts and the ensuing billing of satisfied customers.

In financial services, PwC has garnered enormous amounts of data from the US Census Bureau, US financial data, and other public licensed sources to create secure, a large-scale model of 320 million US consumers’ financial decisions (https://www.pwc.com/gx/en/services/advisory/consulting/security-cyber-assets.html). The model is intended to assist financial services firms map buyers’ personas, simulate “future selves” and anticipate customer behavior. It has empowered these financial services companies in substantiating real-time business decisions within seconds.

Similar to the financial services sector, Artificial Intelligence has been implemented to develop a model of the automobile ecosystem. Here, you have bots that map the decisions made from automotive players, such as vehicle purchasers, manufacturers, and transportation services providers. This has assisted organizations to predict the adoption of electric and driverless vehicles, and the enactment of non-restrictive pricing schemes that work on their target market. It has also assisted them in providing enhanced advertising decision choices.

Artificial Intelligence application development has delivered to marketeers with new and more reliable tools of market forecasting, process automation and decision-making (https://www.tenfold.com/business/artificial-intelligence-business-decisions). Further, Artificial Intelligence can be employed while individuals are scrolling through their social media newsfeed, an algorithm somewhere is determining someone’s medical diagnoses, their parole eligibility, investment and purchasing habits, or their career prospects.

The remaining chapter sections deal with (1) Artificial Intelligence categories, (2) six Throughput Model algorithms driving Artificial Intelligence, (3) impact of Artificial Intelligence, (4) expert systems and Artificial Intelligence systems (including machine learning, neural networks, deep learning and natural language systems).

Categories of Artificial Intelligence

According to Arend Hintze, an assistant professor of Integrative Biology and Computer Science and Engineering at Michigan State University, Artificial Intelligence can be encapsulated into four classes, from the kind of Artificial Intelligence methods that are in existence today to systems that are yet to come into being (https://searchenterpriseai.techtarget.com/definition/AI-Artificial-Intelligence).

Artificial Intelligence can be defined as (1) “narrow” Artificial Intelligence, (2) “general” (or strong) Artificial Intelligence, or (3) “super” Artificial Intelligence. Narrow Artificial Intelligence is able to conduct just one particular task. Further, this Artificial Intelligence type can attend to a task in real-time; however, they pull information from a specific data-set. As a result, these systems do not execute outside of the single task that they are designed to perform.

Narrow Artificial Intelligence has only characteristics consistent with cognitive intelligence. These Artificial Intelligence systems engender a cognitive representation of the world and utilize learning based on previous experiences to bring up-to-date future decisions. Most Artificial Intelligence systems implemented by today’s organizations fall into this group. Examples include systems used for fraud detection in financial services, image recognition, or self-driving cars.

Narrow Artificial Intelligence is not conscious, sentient, or driven by emotions the way that individuals are configured to make decisions. Narrow Artificial Intelligence operates within a pre-determined, predefined range, even if it gives the impression to be much more sophisticated than that due to heuristics and biases (to be discussed in later chapters). Every type of machine intelligence that encircles us in the global society is narrow Artificial Intelligence. Examples of narrow Artificial Intelligence include Google Assistant, Google Translate, Siri and other natural language processing tools. These systems lack the self-awareness, consciousness, and genuine intelligence to match human intelligence. That is, they cannot think for themselves. Nonetheless, narrow Artificial Intelligence by itself is a great accomplishment in human innovation and intelligence.

On the other hand, general Artificial Intelligence’s are more sophisticated. They are able to cope with any generalized task asked of it, much like a human being. In other words, general Artificial Intelligence’s represent machines that display human intelligence, that is, they are able to perform any intellectual task that a human being can in terms of decision-making. This is the kind of Artificial Intelligence that we see in military operations and in movies whereby humans interact with machines and operating systems that are conscious, sentient, and driven by emotions and self-awareness.

General Artificial Intelligence has elements from both cognitive as well as emotional intelligence. Moreover, these systems depict cognitive elements, understand human emotions and include them in their decision making. Affectiva, an Artificial Intelligence firm founded by MIT, uses advanced vision systems to recognize emotions such as joy, surprise, and anger at the same level (and frequently better) as people (https://www.affectiva.com/). Organizations can implement such systems to recognize emotions during customer involvements or while recruiting new employees.

Classes one and two are denoted as “narrow” Artificial Intelligence, whereas, classes three and four are designated as “general” Artificial Intelligence, as follows:

Class 1: Reactive machines. An example is Deep Blue, the IBM chess program, which defeated Garry Kasparov in the 1990s. Deep Blue can recognize pieces on the chess board and generate predictions. Nonetheless, it has no memory and cannot implement previous experiences to update future ones. It scrutinizes possible moves, its own as well as its opponents’. It then selects the next strategic position. Deep Blue and Google’s AlphaGo (a Chinese strategy board game) were designed for narrow purposes and cannot simply be employed to another setting.

Class 2: Limited memory. The automotive industry has fostered several Artificial Intelligence applications, from vehicle design to marketing and sales decision-making support. Moreover, Artificial Intelligence has led to the design of driverless automobiles fortified with multiple sensors that learn and identify patterns. This is utilized through add-on safe-drive features that warn drivers of possible collisions and lane departures.

These Artificial Intelligence systems can utilize previous encounters to inform future decisions. Some of the decision-making functions in driverless self-driving automobiles (oftentimes referred to as an autonomous car/driverless car) are designed in this manner. A self-driving car is a vehicle that utilizes a combination of sensors, cameras, radar and Artificial Intelligence in order to travel between destinations without a human operator. To qualify as fully auto...

Table of contents

• Cover
• Title Page
• Copyright Page
• Preface
• Acknowledgements
• Table of Contents
• 1 Introduction to Artificial Intelligence and Biometrics Applications
• 2 Prelude to Artificial Intelligence
• 3 Artificial Intelligence Six Cognitive Driven Algorithms
• 4 Survey of Biometric Tools and Big Data
• 5 Ethical Issues Addressed in Artificial Intelligence
• 6 Cyber Securities Issues Fraud and Corruption
• 7 Artificial Intelligence, Biometrics, and Ethics Examples
• 8 Conclusions
• Index
• About the Author