PORTLAND, Ore. – Microsoft distinguished scientist Eric Horvitz, who also serves the company as co-director of Microsoft Research (Redmond, Wash.), has been celebrating data, by describing how corporations – from banks to hospitals to The Weather Channel – are using Windows servers to make more accurate predictions in commerce, healthcare and weather prediction.
"I believe that the computational sciences are becoming the noble endeavors of our time, because computing enables so many other things from aerospace to healthcare to science to law to government" said Horvitz in an online Q&A.
Horvitz, who was recently elected to the National Academy of Engineering for his work computational decision making, explains that the streams of mega-data coming into companies today can empower them, but only if they know how to capture, store, interpret and leverage the insights contained in such so-called 'Big Data.' "Effective large-scale data analytics for predictive modeling, visualization, and discovery are becoming central for success," said Horvitz.
The shift of business and commerce activities has made it easier for companies to collect data from transactions, events and sensor streams of every kind. Unfortunately, the software analytics necessary to extract intelligence that can be used from the resultant terabytes of data has lagged behind.
Microsoft Research, thinks it may have machine intelligence breakthroughs that can harvest large amounts of data in real-time with predictive models that do everything from predicting the likelihood that a recently discharged patient will need follow-up healthcare for hospitals, to predicting a consumer's intentions even by "reading" gestures and body language sensed by a Microsoft Kinect 3-D sensor.
Horvitz said that smart personal assistants are coming that will combine machine learning with speech recognition, natural language understanding, machine vision and automated decision making, to solve Big Data problems in all areas including health, education, science, and commerce.
Prediction is the next big thing. It sounds like Minority Report. Well! If the application is elsewhere like Healthcare, it sounds less bad. The challenge is, for example, what are the information patterns needed to call back a recently discharged patient.
There is no doubt when information is so abundant and when computer becomes so powerful. There've got to be some way to make some sense out of the enormous amount of information, including predicting the future.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.