Award-Winning Research Paper Brings Precision to Sampling Methods Used in Statistics and ML

Through a new, efficient algorithm for exact sampling, Microsoft researchers Daniel Tarlow and Tom Minka, along with former Microsoft intern Chris Maddison, address a core problem of statistics and machine learning (ML).

The authors submitted their algorithm at NIPS 2014 where it was picked as one of the two Outstanding Paper Award winners from a field of 1,700 submissions.

Daniel and Tom

“This research makes a very significant advance in the efficiency of sampling, which is a core component of probabilistic modelling and reasoning systems,” said Andrew Blake, Distinguished Scientist and Laboratory Director of Microsoft Research Cambridge.

Tarlow’s hope for the future is these findings will lead to probabilistic reasoning systems that are more powerful and easier to use than current systems. “When we can provide stronger guarantees about the quality of outputs from our inference algorithms, it becomes easier to use these algorithms inside larger systems and to build tools that can be used reliably by non-experts,” he said.

You can read their paper, titled A* Sampling from this site.

ML is a key focus of Microsoft Research and has led to numerous product contributions including Microsoft Office, SQL Server, Xbox One, Cortana speech recognition, and Skype Translator. Additionally, several-state-of-the art ML algorithms from Microsoft Research are offered as part of Microsoft's cloud-based Azure ML platform for predictive analytics.

ML Blog Team