The University of Washington has landed four new faculty members considered among the brightest in the world of computer science.

Share story

They’re a kind of dream team of the computer-science world: Four of the brightest academics in the fields of “big data” and machine learning have been wooed away from top schools to join the University of Washington over the next year.

It’s part of a push by the UW to develop expertise in a field of computer science that is already changing the way people use technology. Their work has the potential to greatly expand the ability to make sense of reams of data, gain a better understanding of the world at large, and make technology more useful in everyday life.

The UW has had “some stunning recruiting successes,” said Peter Lee, corporate vice president of Microsoft Research. The new hires “are all very well-known. Let’s just say, in academic circles, this has made quite a buzz.”

The academics include Carlos Guestrin, a professor at Carnegie Mellon University; Emily Fox and Ben Taskar, both professors at the University of Pennsylvania; and Jeff Heer, a professor at Stanford University.

All four will teach both undergraduate and graduate courses, and do research.

Together, they’re what the university calls a “hiring cluster,” and they beef up the UW’s expertise in what’s broadly known as computational thinking, said UW provost Ana Mari Cauce.

Guestrin and Taskar are both experts in machine learning — building computational systems that improve, and learn, with experience.

A good example of machine learning, Guestrin said, is a spam filter that learns which emails you consider to be spam and becomes better able to weed out unwanted messages.

Another example: a search engine that learns over time which choices are most relevant to you and gives you what you want to know, even when you enter a vague request. Microsoft’s Kinect, the system that allows you to control a game console with body movements, was developed with the help of machine learning, said Lee, of Microsoft.

Systems that detect credit-card fraud rely on machine learning to flag anomalies in a customer’s credit-card use. Increasingly, hospitals are using it to help track patients’ conditions and spot potential problems. “The sky’s the limit,” Lee said.

The university was able to woo the four to Seattle because “the UW presents enormous opportunities for collaboration — in sciences, in medicine, in global health,” said Hank Levy, chairman of the UW’s computer-science department. “And the region has one of the most vibrant high-tech economies. When you put this all together with the Northwest lifestyle, it’s an incredibly attractive story for recruiting.”

Stanford University professor Daphne Koller, an expert on artificial intelligence, had both Guestrin and Taskar as students, and “they are both awesome,” she wrote via email. “UW is to be congratulated on making such great acquisitions, which I believe will propel them into a leadership role (with a select number of other top institutions) in the areas of machine learning and artificial intelligence.”

Both Guestrin and Fox, who are married, were being courted by other universities this year, including MIT and Stanford, said Ed Lazowska, who holds the Bill & Melinda Gates Chair in Computer Science and Engineering at the UW.

To sweeten the deal for Guestrin and Fox, Amazon founder and Chief Executive Jeff Bezos met with them, then established two, $1 million endowed professorships in machine learning, helping to fund their salaries.

The university was also able to afford the new hires because the computer-science department had saved money from previous years, when it did not do any hiring, and Cauce added some money from her discretionary funds.

In addition, during the last legislative session, lawmakers required the UW to shift $3.8 million in funding to engineering — of which computer science is a part — to increase the number of students getting degrees. The university needed to increase the size of the faculty to accommodate more students, Cauce said.

Seattle’s high-tech companies have an especially close relationship with the UW’s computer-science department, not only because they want the university to produce graduates for future jobs, but because the industry often collaborates with the UW on research projects.

Machine learning is also likely to play an important role in scientific advancements — improving disease research, or helping scientists monitor the health of the environment, Guestrin said.

One of his projects at Carnegie Mellon helped find the best locations for sensors that detect contaminants in a city’s water pipes, and another pinpointed the best locations to install sensors for detecting algal blooms in lakes and rivers.

“It’s about thinking about what information you gather from the environment to make the best possible predictions,” he said.

Machine learning and “big data” — data sets so huge that the human mind needs help understanding them — are interrelated, Guestrin said, because machine learning can be employed to help understand data. And statistics, Fox’s expertise, is at the core of both.

Taskar’s expertise is in computational linguistics, or speech recognition, the technology used in cellphones and computers that allows people to talk to their devices and be understood. And he’s also working with computer vision — teaching a computer to recognize an image, such as a face.

Using the two technologies, researchers may one day be able to build robots that can be sent into dangerous situations — such as an earthquake-damaged building — and be commanded to carry out a rescue operation.

“Imagine you could talk to it — it could collaborate with you,” Taskar said. “You could ask it, ‘What do you see?’ and it could report back in some intelligent way.”

On a consumer level, the two technologies might be used to turn your cellphone into a device that can “see” what’s around it, using its camera as an eye, and then process your spoken requests in the context of what it sees, Taskar said.

Both Taskar and Guestrin said they were attracted to the UW because researchers and students work with each other across different departments and disciplines.

“It’s a very collaborative environment, with lots of people working together,” Taskar said. “That doesn’t really happen as much at other places.”

Katherine Long: 206-464-2219 or On Twitter @katherinelong.