In Part One of this article, I explained why business leaders should get excited about machine learning and how it can help improve their operations and decision-making. In this article, I will discuss some of the groundwork that must be done and some of the challenges that must be overcome to make machine learning truly useful for businesses. Shou-De Lin, Chief Machine Learning (ML) Scientist at Appier, asserts, “The right planning and application of machine learning can help businesses grow, compete and prepare for the future.”[1] He adds, “Businesses today are dealing with huge amounts of data and it’s arriving faster than ever before. At the same time, the competitive landscape is changing rapidly and it’s critical to be able to make decisions fast.” Cognitive technologies, like machine learning, can help with corporate decision-making. Bain analysts, Michael C. Mankins and Lori Sherer (@lorisherer), explain, “The best way to understand any company’s operations is to view them as a series of decisions.”[2] They add, “We know from extensive research that decisions matter — a lot. Companies that make better decisions, make them faster and execute them more effectively than rivals nearly always turn in better financial performance. Not surprisingly, companies that employ advanced analytics to improve decision making and execution have the results to show for it.”
The importance of clock-speed in business
Mac McGary (@macmcgary), Executive Vice President of Global Sales at Logility, writes, “For companies, clock-speed can be used to guide dynamic business strategy. The clock-speed framework recommends that you master the ability to quickly and continually design, assemble, fortify and sustain chains of ‘competencies’ that deliver value to your customers.”[3] Lin agrees that the speed of business is crucial. He writes, “As Jason Jennings and Laurence Haughton put it, ‘It’s not the big that eat the small … It’s the fast that eat the slow.’ Business success comes from making fast decisions using the best possible information.” He adds, “Machine learning is powering that evolution. Whether a business is trying to make recommendations to customers, hone its manufacturing processes or anticipate changes to a market, ML can assist by processing large volumes of data to better support companies as they seek a competitive advantage.” As noted in Part One of this article, Tech writer Grace Frenson (@GraceFrenson) suggests real-time decision-making is one of the benefits provided by machine learning. She explains, “Your business can make more informed decisions with machine learning since it can process massive amounts of data in a short amount of time.”[4]
Laying the groundwork and overcoming challenges
Lin notes, “While machine learning offers great opportunities, there are some challenges. ML systems rely on lots of data and the ability to execute complex computations. External factors, such as shifting customer expectations or unexpected market fluctuations, mean ML models need to be monitored and maintained.” The reason models must be monitored and maintained is that ML is not very good at predicting the future if things change. Alan J. Porter (@alanjporter), Head of Strategic Services for [A], explains, “ML is great at recognizing patterns but not much else” and “ML assumes tomorrow is going to be the same as today.”[5] Lin also notes, “In addition, there are several practical issues in machine learning that need to be solved.” Those issues include:
1. Data quality. Lin notes, “Data can be noisy, filled with unwanted information that can mislead a machine learning model into making incorrect predictions.” Shreya Shankar (@sh_reya), a machine learning engineer at Viaduct, asserts, “Data is actually more important than models.”[6] Melanie Chan, a Publishing Executive at Unleashed Software, bluntly states, “Bad data is bad business.”[7]
2. The complexity and quality trade-off. Lin asserts, “Building robust machine learning models requires substantial computational resources to process the features and labels. Coding a complex model requires significant effort from data scientists and software engineers. Complex models can require substantial computing power to execute and can take longer to derive a usable result. This represents a trade-off for businesses. They can choose a faster response but a potentially less accurate outcome. Or they can accept a slower response but receive a more accurate result from the model. But these compromises aren’t all bad news. The decision of whether to go for a higher cost and more accurate model over a faster response comes down to the use case.”
3. Sampling bias in data. Bias is receiving a great deal of attention in today’s media. As a result, any company turning a blind eye to sampling bias could risk a big hit to its reputation. Lin writes, “The problem here isn’t the model specifically. The problem is that the data used to train the model comes with its own biases. However, when we know the data is biased, there are ways to debias or to reduce the weighting given to that data. The first challenge is determining if there is inherent bias in the data.”
4. Changing expectations and concept drift. Lin notes, “Machine learning models operate within specific contexts.” Unfortunately, he notes, “The ML model can drift away from what it was designed to deliver. Models can decay for a number of reasons. Drift can occur when new data is introduced to the model. This is called data drift. It can also occur when our interpretation of the data changes. This is concept drift. … That means you need to keep checking the model.” Shankar notes there are two related concepts companies should be concerned about: Reproducibility and replicability. She explains, “Reproducibility [means]: Given the code and data, can I get the same results as advertised in the paper or technical report? Replicability [means]: If I apply this algorithm on different data, will it still work? Will I get similar results? It’s very common in industry that there are different distributions in the training, validation, and test sets.”
5. Monitoring and maintenance. According to Lin, “Creating a model is easy. Building a model can be automatic. However, maintaining and updating the models requires a plan and resources.” Choosing the right model for the right problem and dataset is also a challenge. At Enterra Solutions®, we leverage the Representational Learning Machine™ (RLM) created by Massive Dynamics™. The RLM can help determine what type of analysis is best-suited for the data involved in a high-dimensional environment. Obtain the right data, identify a specific problem, and apply the right analytics model and you get the results for which you are looking.
Concluding thoughts
Lin concludes, “Machine learning offers significant benefits to businesses. The ability to predict future outcomes to anticipate and influence customer behavior and to support business operations are substantial. However, ML also brings challenges to businesses. By recognizing these challenges and developing strategies to address them, companies can ensure they are prepared and equipped to handle them and get the most out of machine learning technology.”
Footnotes
[1] Shou-De Lin, “Five practical issues in machine learning and the business implications,” ITProPortal, 20 July 2020.
[2] Michael C. Mankins and Lori Sherer, “Creating value through advanced analytics,” Bain Brief, 11 February 2015.
[3] Mac McGary, “Achieve Results Faster by Increasing Your Company’s Clockspeed,” Logility Blog, 14 July 2020.
[4] Grace Frenson, “8 Ways Your Business Can Benefit From Machine Learning,” Market Scale, 8 June 2020.
[5] Alan J. Porter, “Machine Learning Isn’t Rocket Science,” CMS Wire, 9 September 2019.
[6] Shreya Shankar, “Reflecting on a year of making machine learning actually useful,” Shreya Shankar blog, 29 June 2020.
[7] Melanie Chan, “Telltale Signs You Are Using Bad Data,” Unleashed Software, 1 July 2019.
