13 Ways Machine Learning Can Steer You Wrong - InformationWeek

InformationWeek is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them.Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Software // Productivity/Collaboration Apps
07:06 AM
Lisa Morgan
Lisa Morgan
Connect Directly

13 Ways Machine Learning Can Steer You Wrong

Applying machine learning and artificial intelligence to your decision-making can help your business stay competitive. But a lot can go wrong along the way. Without the proper checks and balances, machine learning efforts can spiral out of control, exposing your organization to risks. Here are 13 pitfalls to avoid.
1 of 14

(Image: geralt via Pixabay)

(Image: geralt via Pixabay)

Succeeding in today's fast-paced business economy requires companies to harness data quickly and at scale. As the volume, velocity, and variety of data increase, it's becoming necessary to use machine learning and artificial intelligence (AI) to sift through all the incoming information, make sense of it, and accurately predict future business direction.

Getting machine learning right isn't an easy task, however. It takes the right expertise, the right tools, and the right data to achieve the promise of machine learning. Even with all of those factors in place, it's still easy to get it wrong.

"Machine learning gives us a very powerful set of techniques for making predictions, but it can also lead to disastrous results if you don't understand what your machine learning algorithm is doing," said Spencer Greenberg, a mathematician and founder of decision-making website ClearerThinking.org, in an interview. "It is critical to study the algorithm once it has been trained, to understand how it is making its predictions, and whether what it is doing makes sense from a business perspective."

[Even the best intentions can go horribly wrong. Read 10 Ways to Doom Your Next Mobile App.]

Machine learning is sometimes viewed as a panacea to all business challenges. By failing to consider its realistic potential -- and serious limitations -- it's easy for anyone to misunderstand and misapply machine learning.

"[The] stream of publicity around machine learning milestones reached [by] the big players -- beating human opponents at board games, breakthroughs in medical screening and so on -- gives the impression of continuous, rapid progress, and underplays the frustrations and dead ends," said Monty Barlow, director of Machine Learning at global product development and technology consulting firm Cambridge Consultants, in an interview. "In practice, return on investment [from machine learning] can be late or never."

Organizations are using machine learning in tactical and strategic ways, such as making product recommendations or informing strategic decisions. While the risks of making an irrelevant product recommendation are relatively low, making consistently irrelevant recommendations may fuel customer churn.

Making an ill-informed strategic decision as a result of badly applied machine learning can have serious consequences for your company.

PricewaterhouseCoopers (PwC) has endeavored to quantify the financial impact of big, strategic, data-driven decisions in terms of dollar and shareholder value. It has also attempted to understand the degree to which different industry sectors rely on machine intelligence.

"[In 2014], we asked people to quantify the decisions they're making, and it was in the hundreds of millions of dollars in terms of impact," said Dan DiFilippo, global and US Data & Analytics leader at PwC, in an interview.

"This time, we asked them to express it terms of shareholder value, and most people selected 5% – 50% [of shareholder value]. Even on the low side, 5% of the value of a big organization is something that will move the needle."

Suffice to say, there's a lot at stake. No matter where you are on the path to applying machine learning, it's worth knowing the pitfalls to avoid.

We've identified 13 key ways in which the misapplication of machine learning and artificial intelligence can lead your business astray. Once you've reviewed these, tell us about your own experiences. We'd love to hear from you in the comments section below.

Lisa Morgan is a freelance writer who covers big data and BI for InformationWeek. She has contributed articles, reports, and other types of content to various publications and sites ranging from SD Times to the Economist Intelligent Unit. Frequent areas of coverage include ... View Full Bio

We welcome your comments on this topic on our social media channels, or [contact us directly] with questions about the site.
1 of 14
Comment  | 
Print  | 
More Insights
Newest First  |  Oldest First  |  Threaded View
User Rank: Apprentice
8/18/2016 | 10:17:55 AM
It is cheaper to do ML wrong than it is to do it right
The common thread I am seeing in each of the 13 is that avoiding bad ML results is going to require a greater investment in time, people, or follow-up to ensure that the outcomes of applying ML remain positive. Too many of my clients have a habit of valuing low cost over high quality.

One of the critical steps in QA for any new system or feature is the comparison of actual results to expected results. If the corporate objectives of ML is to find the unexpected insight, how will incorrect unexpected ML outcomes be sifted out from the set of all possible unexpected outcomes?
InformationWeek Is Getting an Upgrade!

Find out more about our plans to improve the look, functionality, and performance of the InformationWeek site in the coming months.

Remote Work Tops SF, NYC for Most High-Paying Job Openings
Jessica Davis, Senior Editor, Enterprise Apps,  7/20/2021
Blockchain Gets Real Across Industries
Lisa Morgan, Freelance Writer,  7/22/2021
Seeking a Competitive Edge vs. Chasing Savings in the Cloud
Joao-Pierre S. Ruth, Senior Writer,  7/19/2021
White Papers
Register for InformationWeek Newsletters
Current Issue
Monitoring Critical Cloud Workloads Report
In this report, our experts will discuss how to advance your ability to monitor critical workloads as they move about the various cloud platforms in your company.
Flash Poll