Re-posted from the Microsoft Research blog.
The 30th annual Conference on Neural Information Processing Systems (NIPS) is on next week, from December 5th through 10th, in Barcelona.
For those of you who are not familiar with NIPS, it’s an annual machine learning and computational neuroscience conference, featuring many exciting talks, demos, poster presentations and workshops, as well as numerous opportunities to connect with peers from academia and industry. The conference has grown rapidly in recent times, with this years’ event expecting a turnout of over 4,500 attendees.
Microsoft will have a strong presence at NIPS this year, as always. Aside from contributing a tutorial co-chair and invited speakers, we will also be delivering several workshops and over 20 poster presentations at the event. You can browse our NIPS participants’ profiles and their contributions to the conference in this post.
In addition, there’s also an interesting post on NIPS by Christopher Bishop, Distinguished Scientist and Director of the Microsoft Research Labs in Cambridge (UK), where he asserts that real substance, not hype, is behind the palpable excitement around AI at NIPS this year. Computing, Chris feels, is undergoing a profound shift from hand-crafted software to solutions that are learned from data. Associated with this shift is a rethink of computation as something that fundamentally involves uncertainty, expressed as probabilities, as opposed to cold hard logic and 0s or 1s. From a probabilistic perspective, machine learning can be viewed as a reduction in uncertainty as a result of observing data – an iterative open-ended process, whereby observations made thus far become “prior inputs” into the next round of decision making, which will be made on the basis of data that will be observed in the future.
We hope to connect with several of you at NIPS next week – safe travels!
CIML Blog Team