Page 16 - Security Today, March 2019

LTooking Beyond the Hype By Sean Lawlor
he past few years have seen significant advancements unrealistic expectations for those considering the technology.
in computing power. With this, machines seem to have Today’s examples of what many consider to be AI in our lives— a greater ability to learn about us and participate in Deep Blue beating a top chess player, Siri recognizing a song, Ama- our lives. Whether through product purchase sugges- zon suggesting a new book—are really examples of increasingly tions on Amazon.com and other retail outlets or in our small computers running a series of algorithms, searching through
business and professional pursuits, machines are busy learning every- huge databases, or doing a lot of calculations very quickly. With their
where around us. Recently, the market has become flooded with buzz words relating to this type of work.
Learning the Proper Terms
Artificial Intelligence, Machine Learning, and Deep Learning are often used inaccurately and interchangeably. Given the significant advancements that have been made in this field, especially in the physical security industry, it is important that we be clear about these terms and their application. Using the term AI loosely only serves to misrepresent what machine learning can do and has the potential to generate misguided and unrealistic expectations.
Artificial Intelligence. AI is a broad term that first appeared in published research in 1956. For years, we understood AI as it ap- peared in pop culture, which lead to questions of a robot’s emotional capacity or their ability to take over the world. AI denotes a fully functional artificial brain that can reason, evolve, self-learn, and make human-like decisions. Currently, we are many years (or decades) away from this. Using the term artificial intelligence (or AI) related to tech- nology or applications today can be inaccurate and potentially raise
faster computing power and processing speeds, our current machines are able to comb through a huge amount of data to provide deeper insights. These results can be more accurately categorized as guesses that can help us make decisions more quickly and efficiently.
Machine Learning. ML is an area of artificial intelligence that uses data to help a computer improve performance without being ex- plicitly programmed. Static programming provides a computer with a set of instructions that do not change over time. Machine learning allows programmers to enable a computer to assess and alter its com- putational processes through training. Specifically, a computer is pro- grammed with algorithms that enable it to determine which features of an input it should use in the identification process to efficiently produce the most accurate output. In a simple example, a computer might be trained to determine whether color or shape is a better indi- cator for correctly classifying a new input.
Working primarily with data in the form of language, text, video, or images, machine learning uses statistical techniques to enable com- puter systems to solve problems, make decisions and predictions, or improve the efficiency of specific, narrowly defined tasks.
14
0319 | SECURITY TODAY
DEEP LEARNING
buffaloboy/Shutterstock.com