“A.I. is a big fat lie,” writes Eric Siegel (@predictanalytic), a former computer science professor at Columbia University. “Artificial intelligence is a fraudulent hoax — or in the best cases it’s a hyped-up buzzword that confuses and deceives. The much better, precise term would instead usually be machine learning — which is genuinely powerful and everyone oughta be excited about it.”[1] Business executives would probably be a lot more excited if they understood a little more about how machines learn. Mark Troester (@mtroester), vice president of strategy at Progress, explains, “Before businesses start to develop a strategy around machine learning and AI, it’s important to review how machines really learn, and how this can impact your AI and machine learning strategies.”[2] In the following paragraphs, I’ll review some of the ways experts have tried to explain machine learning to non-technical audiences.
How Machines Learn
Acquiring data. The need for data should be obvious. Without data there is no machine learning. Business and management journalist Sara Brown (@SaraMarieBrown) explains, “Machine learning starts with data — numbers, photos, or text, like bank transactions, pictures of people or even bakery items, repair records, time series data from sensors, or sales reports. The data is gathered and prepared to be used as training data, or the information the machine learning model will be trained on. The more data, the better the program.”[3] Machine Learning Engineer Aurangazeeb A K discusses the relationship between data and machine learning in a more philosophical way. He writes, “Every data point counts. As you are on a journey of learning you pick each of these stones (i.e., each data point) along your way and get some understanding from it — you learn something from it. As you progress, you will see and pick many stones along the way and continue accumulating knowledge about it. After you have traveled so far, you would have considerable amount of understanding accumulated about all those stones (or data points).”[4] He cites a speech by Steve Jobs in which Jobs concluded, “You cannot connect the dots by looking forward but only by looking backward.” Software engineer Ben Dickson (@bendee983) dumbs it down a little further by explaining that data sets are really nothing more than “examples.” He writes, “Machine-learning AI models are developed not by writing rules but by gathering examples. … Machine learning is applicable to many real-world tasks, including image classification, voice recognition, content recommendation, fraud detection, and natural language processing.”[5]
Looking for patterns. The most common words found in machine learning explanations are “patterns” and “algorithms.” José Luis Espinoza, a data scientist at BBVA Mexico observes, “Ultimately, machine learning is a master at pattern recognition, and is able to convert a data sample into a computer program that can extract interferences from new data sets it has not been previously trained for.”[6] Oliver Tan, Co-Founder and CEO of ViSenze, writes, “Machine learning is a form of artificial intelligence, and if you ask a computer scientist, you will get a highly technical answer that involves algorithms, pixels and both supervised and unsupervised learning. In simpler terms, a machine ‘learns’ by looking for patterns among massive data loads, and when it sees one, it adjusts the program to reflect the ‘truth’ of what it found.”[7] Tech writer Cal Jeffrey (@CSJeffrey) combined machine learning definitions from MIT, Stanford, and Carnegie Mellon into this explanation: “Machine learning involves training a computer with a massive number of examples to autonomously make logical decisions based on a limited amount of data as input and to improve that process with use.”[8] The following video explains how the Pew Research Center uses machine learning.
Applying algorithms. The magic behind machine learning is created by algorithms. Data Iku created the following infographic that provides a great overview of machine learning algorithms and how they are used.
Leveraging mathematics and statistics. Business leaders don’t necessarily need to understand the math behind many machine learning techniques; however, the experts running your system should. Mathematician Richard Han explains, “Machine learning is a wildly popular field of technology that is being used by data scientists around the globe. Mastering machine learning can be achieved via many avenues of study, but one arguably necessary ingredient to success is a fundamental understanding of the mathematics behind the algorithms. Some data scientists-in-training often try to take a shortcut and bypass the math, but that route is shortsighted. In order to get the most out of machine learning, you really need that important perspective for what the algorithm is really doing behind the scenes. This perspective is only available with the math.”[9] If you want to learn more, download Han’s paper by clicking on this link.
Although math is certainly an integral part of machine learning, data scientist Niels Goet (@NielsGoet) explains there are philosophical differences between how machine learning uses math to create models and how fields like statistics use math to create traditional statistics modeling. He writes, “Machine learning and traditional statistics modeling have fundamentally different philosophical roots. … Machine learning’s main purpose is to ‘learn’ from data, that is, to optimize the weights on features (or: parameters), and to apply that information to generate new predictions. … By contrast, traditional statistical modeling attempts to use mathematical formulas to formalize the relationship between two or more variables.”[10] Knowing differences between machine modeling and traditional statistics modeling, Goet explains, is important because they accomplish different things.
“Your choice between either,” he writes, “should really be informed by the data problem that you’re facing. For example, traditional statistics modeling is preferable by far if you have a relatively small dataset that consists of structured data, and if your purpose is to identify if and to what extent a (limited) set of variables affect a phenomenon of interest. In other words: use the traditional statistics modeling toolbox when you are interested in taking a deductive approach to your research problem, i.e., when you start from theory and hypotheses, and use empirical data to confirm or reject your theoretical propositions. By contrast, machine learning is your go-to strategy if you are dealing with large volumes of unstructured data, and your goal is to predict the extent or occurrence of a phenomenon as accurately as possible.”
There are also differences between data analytics and machine learning. Vance Reavie (@VanceReavie), CEO and Founder at Junction AI, explains, “Machine learning is a continuation of the concepts around predictive analytics, with one key difference: The AI system is able to make assumptions, test and learn autonomously. … Data analysis refers to reviewing data from past events for patterns. Predictive analytics is making assumptions and testing based on past data to predict future what/ifs. [And] machine learning analyzes data, makes assumptions, learns and provides predictions at a scale and depth of detail impossible for individual human analysts.”[11]
Concluding Thoughts
Brown concludes, “With the growing ubiquity of machine learning, everyone in business is likely to encounter it and will need some working knowledge about this field. A 2020 Deloitte survey found that 67% of companies are using machine learning, and 97% are using or planning to use it in the [future].” Jeffrey adds, “Machine learning and deep learning algorithms have infinite room for growth, and we’re sure to see even more practical applications entering the consumer and enterprise markets in the coming decade.”
Footnotes
[1] Eric Siegel, “Why A.I. is a big fat lie,” Big Think, 23 January 2019.
[2] Mark Troester, “Demystifying machine learning: How do machines really learn?” IoT Agenda, 30 January 2018.
[3] Sara Brown, “Machine learning, explained,” MIT Sloan Management School, 21 April 2021.
[4] Aurangazeeb A K, “What is Learning in Machine Learning?” Towards Data Science, 10 November 2019.
[5] Ben Dickson, “What Is Machine Learning?” PC Magazine, 8 July 2019.
[6] Staff, “Machine learning: What is it and how does it work?” BBVA, 8 November 2019.
[7] Oliver Tan, “How Does A Machine Learn?” Forbes, 2 May 2017.
[8] Cal Jeffrey, “Explainer: What Is Machine Learning?” Techspot, 8 July 2020.
[9] Richard Han, “The Math Behind Machine Learning,” insideBIGDATA, 8 November 2018.
[10] Niels Goet, “When Does Traditional Statistics Become Machine Learning?” OXPOL, 31 July 2018.
[11] Vance Reavie, “Do You Know The Difference Between Data Analytics And AI Machine Learning?” Forbes, 1 August 2018.
