According to research conducted by Vanson Bourne for SoftServe, “62% of medium to large organizations expect to implement machine learning for Business Analytics purposes within the next two years.”[1] A press release concerning SoftServe’s Big Data Survey also revealed, “Big Data Analytics technology, despite being relatively nascent, is already widespread with 86% of organisations having systems in place. Furthermore, medium to large organizations see it as a necessity and are receptive to new technologies that build upon Big Data Analytics. Respondents were asked what they saw as the biggest area of opportunity for Big Data in comparison to traditional systems, with 62% agreeing that they consider real time analysis as the biggest area of opportunity today.” Whenever you hear or read about “real-time” or “near-real-time” analysis, you can be assured that artificial intelligence (AI) and machine learning (most likely in the form of a cognitive computing system) are working in the background. I can say that with confidence since it is impossible for human analysts manually to pore through the oceans of data being gathered by today’s companies in a timely fashion. And since data is gathered around the clock, companies need analytics around the clock as well. In that regard, companies are finding that machine learning solutions are their most tireless employees. Graham Templeton (@GrahamTempleton) observes that another reason companies are opting to implement AI solutions is “the ability of modern artificial intelligence to adapt to just about any problem, and improve on the best mankind has managed on its own.”[2]

What is Machine Learning?

Todd Jaquith (@toddmjaquith) writes, “We now live in an age where machines can teach themselves without human intervention. This perpetual self-education can produce insights that are helpful in making proper and productive decisions for us across a variety of fields, from medicine to interstellar space travel.”[3] He goes to note, “Machine Learning (ML) deals with systems and algorithms that can learn from various data and make predictions. An example is predicting traffic patterns at a busy intersection — a program can run a machine learning algorithm containing data about past traffic patterns and, having ‘learned’ previous data, it can devise better predictions of future traffic patterns. … The main goal of a learner is to generalize, and a learning machine able to do that can perform accurately on new or unforeseen tasks. The goal for the learning machine is to mimic human cognition by creating a generalized model to produce precise enough predictions.” From that description, you understand why it’s no coincidence that machine learning has become an important business tool in the Big Data era. Machine Learning systems need lots of data on which to train. Jaquith notes there are several categories of machine learning. They include:

• Supervised ML — A technique that “relies on data where the true label is indicated. Example: teaching a computer to distinguish between pictures of cats and dogs with each image tagged ‘cat’ or ‘dog.’ Labeling is normally performed by humans to guarantee high data quality.”
• Unsupervised ML — This technique “deprives a learning algorithm of the labels used in supervised learning. Usually involves providing the ML algorithm with a large amount of data on every aspect of an object. Example: presented with images of cats and dogs that have not been labeled, unsupervised ML can separate the images into two general groups based on some inherent characteristics of the images.”
• Reinforcement Learning — This technique provides specific instructions to a machine (e.g., rules for playing chess) and is then provide feedback on how well it used those rules to perform a task. “Example: Learning to play chess. ML receives information about whether a game played was won or lost. The program does not have every move in the game tagged as successful or not, but only knows the result of the whole game.”

Jaquith notes there are over a dozen approaches involved in ML including: decision tree learning; artificial neural networks; deep learning; and Bayesian networks.

Humans Should Still Provide Oversight

There are some famous examples of how machine learning can go astray without human intervention. A few years ago, Eric Blattberg (@EricBlattberg) reported, “When deep learning startup AlchemyAPI exposed its natural language processing system to the Internet, it determined that dogs are people because of the way folks talked about their pets. That might ring true to some dog owners, but it’s not accurate in a broader context. That hilarious determination reflects the challenges — and opportunities — inherent to machine learning.”[3] More recently Tay — the Twitter chat bot that Microsoft launched on 23 March 2016 — had to be taken off line after a single day. As Anthony Lydgate (@anthonylydgate) reports, “She became a racist, sexist, trutherist, genocidal maniac.”[4] Tay’s case says a lot more about people who tweet than it does about how machine learning works. In an even more recent case, Facebook got rid of human editors for its trending news offering and turned the job over to machine learning algorithms with less than satisfactory results. Sam Thielman (@samthielman) reports, “The results, so far, are a disaster.”[5] How bad did it get? Thielman reports, “The fully automated Facebook trending module pushed out a false story about Fox News host Megyn Kelly, a controversial piece about a comedian’s four-letter word attack on rightwing pundit Ann Coulter, and links to an article about a video of a man masturbating with a McDonald’s chicken sandwich.” The above examples demonstrate that machines still need help “getting it right.”

Benefits to Business

Fortunately, getting it right in business is generally easier than turning a machine loose to learn from the Internet. Mike Feldner, regional head at Mu Sigma, notes, “Machine learning technologies leverage enormous amounts of data to help make predictions that go way beyond the capabilities of manual processing. This results in a huge boost to workplace efficiency while decreasing the risk of human error.”[6] He suggests there are a few areas where machine learning can most improve workplace operations. They are: Customer service and retention; business operations; and cyber security. Ronald van Loon (@Ronald_vanLoon) agrees with Feldner that machine learning has place in the business world. “With machine learning,” he writes, “the concepts of big data and fast data analytics can be used in combination with artificial intelligence (AI) to avoid … problems and challenges in the first place.”[7] He goes on to report, “According to a recent study by Bain, companies that use machine learning and analytics are: Twice as likely to make data-driven decisions; five times as likely to make decisions faster than competitors; three times as likely to have faster execution on those decisions; and twice as likely to have top-quartile financial results.”

Summary

Feldner concludes, “We are merely beginning to tap into the power of AI in the workplace. New ways to leverage AI technologies in the workplace will continue to be developed and optimized through the coming years. As we create the right balance between humans and machines, we will find better ways to manage and grow our businesses.” Templeton adds, “The fruits of applying truly advanced machine learning, from fully reactive AI voice interfaces to decreased power bills, will change the corporate world in a big, noticeable way.”

Footnotes
[1] “62% of Organizations Expect to Implement Machine Learning to Big Data by 2018,” PR Newswire, 16 June 2016.
[2] Graham Templeton, “Machine learning is about to change how corporations are run,” ExtremeTech, 22 July 2016.
[3] Todd Jaquith, “Understanding Machine Learning Infographic,” E-learning Infographics, 22 July 2016.
[3] Eric Blattberg, “Cognitive computing is smashing our conception of ‘ground truth’,” Venture Beat, 20 March 2014.
[4] Anthony Lydgate, “I’ve Seen the Greatest A.I. Minds of My Generation Destroyed by Twitter,” The New Yorker, 25 March 2016.
[5] Sam Thielman, “Facebook fires trending team, and algorithm without humans goes crazy,” The Guardian, 29 August 2016.
[6] Mike Feldner, “Machine Learning and AI in the Workplace: The Future of Business Tools,” Information Management, 26 July 2016.
[7] Ronald van Loon, “Machine Learning Becomes Mainstream: How to Increase Your Competitive Advantage,” Datafloq, 22 August 2016.