09 May

The Power of Machine Learning

An Article By, David Tan, Chief Technology Officer

If you’ve read any articles or blogs that I’ve written, you’ve probably figured out what a huge fan I am of technology.  From cool new toys and gadgets to software and systems that change the way we live and work, the ever-changing landscape of this industry keeps me energized every day.  Usually when I write about interesting new technology, I’m relating it to things you can do in your business.  Ways to increase profits or productivity, or even new services you can start offering your customers.  This is going to be a little different.  Today I want to talk about how the advances in technology can benefit us in our lives on a daily basis.

One of the themes I’ve talked and written a lot about recently is Machine Learning and Artificial Intelligence.  People are worried that computers will continue to get smarter and robots will get better and suddenly there will be no jobs left for humans!  Obviously, this isn’t the case.  AI is going to make our world a better place.  Artificial Intelligence isn’t about replacing people, it’s about improving our abilities and finding ways to make us better at what we do.  To show that, I want to talk about an amazing article I read recently.

I’m very fortunate in that I have never been impacted by suicide in my life directly.  I know many people who have and it’s a devastating experience to have someone you care about take their own life.  Nearly 45,000 Americans take their lives every year, 120 a day.  Yet despite the magnitude of this issue, there has been precious little advancement in the study of suicide prevention.  A recent study by a Florida State Psychology Assistant Professor revealed that 50 years of suicide prediction research had not produced any real progress in being able to predict who will try to kill themselves. The traditional risk factors identified over the past half century to predict suicidal behavior — such as depression, stress or substance abuse — could muster an accuracy rate not much better than random guessing.  A coin flip is as accurate on suicide attempts as the brightest suicide experts in the world.

However, there is hope.  Jessica Ribeiro, a psychology researcher at Florida State recently published a paper titled “Predicting Risk of Suicide Attempts over Time through Machine Learning.”  In the paper, it finds that machine learning can predict with 80-90 percent accuracy whether someone will attempt suicide as far off as two years into the future.  The algorithms become even more accurate as a person’s suicide attempt gets closer. For example, the accuracy climbs to 92 percent one week before a suicide attempt when artificial intelligence focuses on general hospital patients.

The data for the research comes from over 2 million anonymous patient electronic health records.  Of the patients, 3200 attempted suicide.  Intelligent machines are able to sort through and understand the data at a rate that is impossible to humans.  Plus, they can pick up the most unperceivable of signals in the data that no person could ever grasp.  Being able to analyze so much data at once allows the computer to create hypotheses.  These are then tested and tweaked and refined over time as more and more data is ingested.  Ultimately, the computer develops a model it can use to compare data in real-time and hopefully, in this case, create some sort of an early warning system for at-risk patients.

You see technology in general and artificial intelligence specifically is not about replacing people.  It’s about finding ways to accomplish things that humans just can’t do on their own.  We have spent decades building better, faster and smarter technology.  It’s wonderful when it can be used to make our world a better place.  We all owe it to ourselves to embrace these advances and find other ways we can use them for good.

Share this

Leave a reply