Machine learning (ML) is an area of computer science that uses data to extract algorithms and learning models. It applies learned generalizations to new situations and tasks, which don’t involve direct human programming. It has also become a major part of big data and analytics practices, helping to identify hidden insights and make smarter recommendations that inform human decision-making.

Although machine learning relates to artificial intelligence (AI), it’s not solely machines acting in “smart,” human-like ways. It isn’t a magic solution to all our data needs, but is incredibly useful, generating powerful outcomes that save time and reduce tedious, sometimes costly tasks.

The essentials: how machine learning works

The power of machine learning is seen as algorithms are paired with data to generate useful, even interesting predictions that businesses, governments, and society can act on to tackle problems—for instance, more accurate customer segmentation, customer lifetime value prediction, improved public policy, and autonomous driving. Machine learning is not like normal computer programming—instructions aren’t explicitly coded to tell the machine what to do; nor is ML like AI because it doesn’t make autonomous decisions.

For individuals with machine learning apprehension, it likely feels like abstract magic, but it’s superior mathematics and a carefully trained process that can benefit us all. Two main methods exist to guide machine learning models and algorithms—known as supervised and unsupervised learning—though there are other, uncommon methods that are used more frequently now, including semi-supervised learning, which combines elements of other methods.

The best learning happens when a machine learning model adapts to or extrapolates data without human intervention, but the combined power of humans and machines learning from data is where the rubber meets the road.

The connection between data and machine learning’s impact on the world

Some skeptics argue against the benefits of machine learning, but the list of advantages and use cases grows as individuals and organizations embrace the benefits it provides. Machine learning is ultimately helping humans do more sophisticated analysis—to get answers faster and uncover what are sometimes hidden, surprising insights. In many cases, it has reduced or eliminated tedious work that once stifled human creativity, required considerable time, and took attention away from strategic, creative, and critical tasks or decision-making.

You may be surprised to learn that machine learning is already all around you—from online product recommendations and email spam filtering to advances in healthcare and finances. For instance, JPMorgan Chase and Verizon use machine learning to enhance their analytics and analyze overwhelming, diverse sets of data, which are too vast or complicated for human review alone.

Be mindful, however, machine learning is only as useful as the data we feed it. We create data every single second—in fact 2.5 quintillion bytes are created daily, according to IBM—and we need to store all of it, too. With better technologies that now help clean, shape, examine, and filter data for more accurate and deeper examination, we can harness and feed it to smart models that then enhance analytic or business intelligence processes. When applied correctly, you’ll find machine learning is unobtrusive, seamlessly working to benefit and solve problems in businesses and society.

Helpful perspectives on machine learning

A quick media scan or online search for machine learning will net a long list of results. So where do you start to get a baseline understanding and enhance your knowledge as you couple it with data analytics? Have a look at this roundup of perspectives from business, technology, and other experts. They’ve been in the trenches examining machine learning and some of the technology solutions that also extract maximum value from data sets, large or small, simple or complicated.

  • A virtual conference series from Google and Tableau features six exceptional thought leaders and their experiences with machine learning—from theory to practical applications in day-to-day business.
  • This byline discusses the delicate balance of using both machine and human intelligence, and recognizing their distinct advantages.
  • Gartner research director Erick Brethenoux explains the five categories of impact from data science and machine learning, and provides real-world examples taken from the worlds of government, business, etc.
  • Follow industry leaders to learn about developments and applications of machine learning and artificial intelligence with this roundup of informative ML and AI blogs.