Machine Learning – the future is now!

Machine Learning – it’s real

After many decades of minimal progress, Machine Learning has finally spawned many successful applications these last few years. This sudden burst in innovation has contributed to the public perception that a major information revolution could be just around the corner, one that is perhaps comparable to the Renaissance.

The future is now

Futurists twenty years ago predicted driverless cars, but it is now actually happening; if you’re skeptical, you should be aware that legislation is moving through Congress to regulate it. “Automated homes” already handle many daily needs like heating and lighting, but in the future your home should have the ability to reliably monitor the health and safety of our elders and toddlers. Hot products are flowing from the Big 5 tech firms that increasingly integrate with many third-party platforms – think Amazon’s Alexa handling all your daily, mundane tasks.

Corporations have already taken advantage

While the consumer applications are getting all the press, there are also massive industrial applications already running that are reducing operating costs and delivering better products. Companies are using algorithms to chill data centers more efficiently, mine product sales data to prepare for next month’s inventory, and monitor products sliding along assembly lines at small factories. Seemingly, any extremely repetitive or complex task with abundant companion data could soon be streamlined using machine learning. It may soon be less natural to ask ‘What will machine learning do?’, rather than ‘what will Machine Learning not be doing?’

Machine Learning and Health Care

But back to the infinite possibilities for benefitting individuals: let us focus on health care. There are dozens of low-cost sensors already being deployed to monitor our bodies’ health, including blood pressure monitors, heart rate sensors, and sensors that can monitor a diabetic’s blood sugar; the next step will be to correlate this sensor data with food intake and activity levels, and help people manage their condition better. The huge datasets generated in this field will give researchers not just a sample of a population to study, but the population itself. It will be exciting to see which applications bloom.

Machine Learning and… …everbeat

For us at everbeat, where our focus is on heart health, we will be using Machine Learning in many ways. Classic techniques like Bayesian analysis will help our customers interpret risk from the heart data generated by the everbeat watch. We will use neural network algorithms to classify and help physicians zero in on potential heart conditions. We believe Machine Learning can help our customers perform a basic cardiac stress test at home, rather than in a hospital or doctor’s office. Finally, combining everbeat’s body sensor data with behavioral data provided by our customer, we hope to be able to provide valuable insights to the customer about what behaviors are better than others at improving the quality of their health, and more broadly their lives. We believe the dual promise of precision medicine and value-based healthcare are right around the corner, and Machine Learning is a crucial ingredient to achieving each future. We at everbeat are planning to help enable it – one heart at a time.


Scott Riccardelli

John McCauley