Tony Hey

<article-ideainbrief class="article-ideainbrief">
 <h4 class="ideainbrief-title">The Idea in Brief</h4>
 <div class="ideainbrief-contents">
  <p>Because of our ability to collect and analyze vast quantities of data, scientists now have the potential to solve some of the world’s biggest problems. But until very recently, we didn’t know how to combine the right data sets and see crucial patterns.</p>
  <p><strong>The turning point:</strong> By utilizing 21st-century computing power, human expertise, and a systematic approach to storing and mining information, scientists are beginning to achieve real breakthroughs. They’re also opening the process up to other experts—and even the public—by making their data transparent and available.</p>
  <p><strong>A hope for the future:</strong> Today researchers at Microsoft and other organizations are using these methods to attack problems in astronomy, oceanography, health care, water management, and climate change. These tools also have the potential to bring profound improvements to business.</p>
 </div>
</article-ideainbrief><p>A visitor walking the halls of Microsoft Research’s campus in Redmond, Washington, today is likely to overhear discussions not only about computer science but about a surprising variety of other subjects, from which way a galaxy rotates, to a new AIDS vaccine, to strategies for managing the planet’s precious supply of fresh water.</p>

wg0BppL1nshci6qidCOlWTKrkhjq2It3bHYJnSTDN8BaMt6Z5HQsWo473czV9gfqA1KKuUkXP931rvgY+ymXYCgexz7EG/G1p8aqydWUs9HQI6aL+ZhltduEuTFkPPHw9RAmpqaYnCJCdaO9I5Y9dw==

<p><span class="reprint">Reprint: R1011B</span></p>
<p>For decades, computer scientists have tried to teach computers to think like human experts. Until recently, most of those efforts have failed to come close to generating the creative insights and solutions that seem to come naturally to the best researchers, doctors, and engineers. But now, Tony Hey, a VP of Microsoft Research, says we&#8217;re witnessing the dawn of a new generation of powerful computer tools that can &#8220;mash up&#8221; vast quantities of data from many sources, analyze them, and help produce revolutionary scientific discoveries.</p>
<p>Hey and his colleagues call this new method of scientific exploration &#8220;machine learning.&#8221; At Microsoft, a team has already used it to innovate a method of predicting with impressive accuracy whether a patient with congestive heart failure who is released from the hospital will be readmitted within 30 days. It was developed by directing a computer program to pore through hundreds of thousands of data points on 300,000 patients and &#8220;learn&#8221; the profiles of patients most likely to be rehospitalized. The economic impact of this prediction tool could be huge: If a hospital understands the likelihood that a patient will &#8220;bounce back,&#8221; it can design programs to keep him stable and save thousands of dollars in health care costs. </p>
<p>Similar efforts to uncover important correlations that could lead to scientific breakthroughs are under way in oceanography, conservation, and AIDS research. And in business, deep data exploration has the potential to unearth critical insights about customers, supply chains, advertising effectiveness, and more.</p>

The Big Idea: The Next Scientific Revolution

Reprint: R1011B For decades, computer scientists have tried to teach computers to think like human experts. Until recently, most of those efforts have failed to come close to generating the creative insights and solutions that seem to come naturally to the best researchers, doctors, and engineers. But now, Tony Hey, a VP of Microsoft Research, says we’re witnessing the dawn of a new generation of powerful computer tools that can “mash up” vast quantities of data from many sources, analyze them, and help produce revolutionary scientific discoveries. Hey and his colleagues call this new method of scientific exploration “machine learning.” At Microsoft, a team has already used it to innovate a method of predicting with impressive accuracy whether a patient with congestive heart failure who is released from the hospital will be readmitted within 30 days. It was developed by directing a computer program to pore through hundreds of thousands of data points on 300,000 patients and “learn” the profiles of patients most likely to be rehospitalized. The economic impact of this prediction tool could be huge: If a hospital understands the likelihood that a patient will “bounce back,” it can design programs to keep him stable and save thousands of dollars in health care costs. Similar efforts to uncover important correlations that could lead to scientific breakthroughs are under way in oceanography, conservation, and AIDS research. And in business, deep data exploration has the potential to unearth critical insights about customers, supply chains, advertising effectiveness, and more.

The Big Idea: The Next Scientific Revolution

Summary.

Partner Center