NYU Data Science newsletter – June 11, 2015

GROUP CURATION: N/A

A few weeks ago, Sonal Chokshi of Andreessen Horowitz and I chatted on an a16z podcast. Here’s her summary of that conversation:

On this episode of the a16z Podcast, Nguyen puts on his former computer science professor hat to describe “Big Data” in relation to “Machine Learning”— as well as what comes next with “Deep Learning”. Finally, the former Google exec shares how Hadoop and Spark evolved from the efforts of companies dealing with massive amounts of real-time information; what we need to make machine learning a property of every application (Why would we even want to?); and how we can make all this intelligence accessible to everyone.

Teaching machines to read natural language documents remains an elusive challenge. Machine reading systems can be tested on their ability to answer questions posed on the contents of documents that they have seen, but until now large scale training and test datasets have been missing for this type of evaluation. In this work we define a new methodology that resolves this bottleneck and provides large scale supervised reading comprehension data. This allows us to develop a class of attention based deep neural networks that learn to read real documents and answer complex questions with minimal prior knowledge of language structure.

The spring 2015 edition of the Machine Learning & Data Science (MLDS) Conference for Microsoft employees concluded two weeks ago. This conference, which is held twice yearly, is a key part of the vibrant community around ML and data science within Microsoft. In it, we cover the entire spectrum of Advanced Analytics (AA), from data discovery and ML algorithms to the end-to-end data science process and real-world customer apps.

These technologies are increasingly used in a new generation of intelligent apps that are bringing about unprecedented insights into our ever-expanding universe of digital data.

The event saw tremendous interest with thousands of attendees, both in-person and employees joining online around the world. The conference featured over 65 talks, 60 posters/demos, and 15 tutorials, each of which shone a bright light on cutting-edge work happening in different parts of our company. Here’s a short video to give you a flavor of this year’s spring event:

Craig McLuckie took the idea to Urs Hölzle, the man who oversees Google’s global network of computer data centers, and Hölzle didn’t like it.

Together with two other engineers in Google’s Seattle office, McLuckie wanted to recreate Borg as an open source project. Borg is the sweeping software tool that drives everything from Google Search to Gmail to Google Maps, letting the company carefully parcel computing tasks across that global network. For years, it was one of the company’s best kept secrets. And McLuckie wanted to share its blueprint—or at least some of it—with the rest of the internet.

“When we went to Urs the first time, he was not really happy about the idea,” McLuckie says. “What we were basically doing is offering up the secret sauce that runs Google as an open source technology.”

Celebrated Swedish statistician Hans Rosling visits the World Bank and tells Pabsy Pabalan that he supports the Bank’s open data policy; believes statistics represent stories about people and that his favorite number is zero. Then he pulls out some dollar bills and things get even more interesting. #PabsyLive

Tesla CEO Elon Musk worries it is “potentially more dangerous than nukes.” Physicist Stephen Hawking warns, “AI could be a big danger in the not-too-distant future.” Fear mongering about AI has also hit the box office in recent films such as Her and Transcendence.

So as an active researcher in the field for over 20 years, and now the CEO of the Allen Institute for Artificial Intelligence, why am I not afraid?

International Business Machines Corp. is pouring resources into a free software product as the giant seeks to gain influence in a fast-growing area of technology.

The Armonk, New York-based company has created a Technology Center in San Francisco to focus on a free open-source software project called Spark, according to IBM executive Rob Thomas. IBM hired 20 people within the last month, Thomas said in a video posted online June 3.

In this book we look at machine learning from a fresh perspective which we call model-based machine learning. This viewpoint helps to address all of these challenges, and makes the process of creating effective machine learning solutions much more systematic. It is applicable to the full spectrum of machine learning techniques and application domains, and will help guide you towards building successful machine learning solutions without requiring that you master the huge literature on machine learning.