One of the most overhyped and underestimated forms of digital technology might be machine learning (ML), and it is shaking up the foundation of the digital business. Data-pattern anomalies are detected in real time to weed out spam, fraud, and potential security threats. People are working more efficiently and strategically as routine tasks are automated. And even the definition of ‘business intelligence’ is evolving from a tool of reflection to a resource for real-time – and forward-looking – truth.
And if Jim Harris, author of Blindsided: How to Spot the Next Breakthrough That Will Change Your Business Forever, is correct, this is only the beginning of machine learning’s influence. “When combined, ML, AI, and cognitive computing revolutionize the way we work, how industries operate, and how society functions,” he acknowledges in the SAP e-book “The Path to Digital Innovation.”
Opening the door to insight-driven digital transformation
In recent years, data growth has brought unprecedented complexity to businesses of all sizes and regions. Everything from employee hiring, retention, and engagement to asset management and product pricing has become an interconnected factor in a massive ecosystem of layers upon layers of data. And, quite frankly, most organizations are struggling to understand how to extract the most value from this information.
Fortunately, the vision of machine learning embedded in technology, processes, and machines is quickly becoming a reality. For businesses that are prepared, tremendous opportunity for growth is on the horizon as their workplace culture becomes naturally data-driven.
In fact, Aberdeen research revealed that organizations with a data-driven culture have achieved a seven percent annual increase in revenue by improving efficiency and accelerating operational cycles. Plus, they decreased functional costs by 12%, which were regarded as wasteful spend before the transition.
Unleashing the full potential of data with groundbreaking computing power
Machine learning may seem like a clear path to solving the jigsaw of data value. However, as Gregory Piatetsky-Shapiro, president and editor of KDnuggets, warns, “[The technology] will improve efficiency, but has a danger of limiting choices since customers are likely to see only what they are predicted to like.”
Recommended for You
Webcast, May 8th: Customer Insights That Drive User Engagement
But as Mike Flannagan, senior vice president of analytics at SAP, argues, the accuracy of those choices boils down to the quality of the data. In an interview with CXO Talk, he observes, “A lot of companies have a garbage-in, garbage-out problem. If you use incorrect data as training data for machine learning algorithms, your predictions are all going to be wrong…and all these advanced technologies, all of these new techniques from learning from data, will not benefit you in any way.”
Nvidia Corporation is working to bring data accountability to machine learning by delivering the power of an entire row of servers within the confines of a single integrated hardware and software supercomputer. “In the world of machine learning, there are two aspects: training and inference,” explains Jim McHugh, vice president and general manager at Nvidia. “The training is when you take all of the data and you teach [the application] to do things like image recognition or the rationalization components of it. Once [the application] learns to a certain extent, then you put it in inference. In essence, [the application] will infer based on what it was trained on.”
By designing a platform with tremendous processing power, Nvidia is enabling businesses to benefit from machine learning capabilities that are continuously learning from the data it receives as well as observed human actions and behaviors. For example, the use of natural language processing and deep learning techniques on Nvidia’s GPU platform enables analysis of unstructured data and creates automated rules to categorize and route tasks to the right process or alert the right people when exceptions or errors arise.
Digital transformation and machine learning: It’s all connected
As companies move deeper into their digital transformation, machine learning is shaping up to become a critical approach to unlocking unprecedented insights and establishing tight integration between their IT systems and business processes. Every functional area – from HR and finance to sales and supply chain – will finally derive the most value from its data to deliver cost-efficient operations, process accountability, and meaningful work. But most importantly, the promise of delighting customers consistently will become a reality.
For more executive insights on leveraging machine learning, download the SAP e-book “The Path to Digital Innovation.”