What Are the Data Science Outcomes from AI? (VIDEO)

Video produced by Steve Nathans-Kelly

In his closing keynote at Data Summit 2019, John O’Brien, principal advisor and chief researcher, Radiant Advisors, discussed organizations need to take in order to get the outcomes they seek from artificial intelligence and machine learning.

"Most companies are very comfortable with machine learning. They are comfortable with seeing the black box aspect of deep learning, but they don't really fully understand neural networks: How do they work? It's like magic to some degree—a lot of math, a lot of science there, but most people are treating it as, hey we train it this way, we get a high probability, very accurate, high confidence level—and it works. And so they're moving to stronger, more complex capabilities in deep learning."

DBTA’s next Data Summit conference will be held May 19-20, 2020, in Boston, with pre-conference workshops on Monday, May 18.

Basically, machine learning breaks into both supervised and unsupervised learning, he said. "Supervised learning means you create datasets which means we have to have data management to produce quality, governed, good unbiased datasets to begin with. That is the biggest challenge, right there. And, then from there, we can either go in to classification or regression routines depending on the problem we're tackling."

From there, he said, there are about four or five different kind of algorithms that can be applied, identifying the best algorithm for the scenario at hand. And then there is unsupervised learning where data is coming in continuously and training the model as it goes. "This could be your live data, your live feedbacks and the unsupervised learning keeps it current."

Machine learning, and most of AI, is only as good as the data used to train it, emphasized O'Brien. "And one of the things that we can look to is what are going to be the outcomes because a big part of our movement right now in the, if you will, the second era as we move in to AI, people are already starting to say, 'Okay, what's the business outcome for this? What can we do better?'"

Many presenters have made their slide decks available on the Data Summit 2019 website at

To access the video of the full presentation, "Bring it Home: How to Advance Your Analytic Strategies," go to