NYU Data Science newsletter – April 21, 2015

GROUP CURATION: N/A

Dr. Andrew Ng is Chief Scientist at Baidu. He leads Baidu Research, which includes the Silicon Valley AI Lab, the Institute of Deep Learning and the Big Data Lab. Dr. Adam Coates is Director of Baidu Research’s Silicon Valley AI Lab.

In 2011, Time magazine reported research that predicted we would be consuming 966 exabytes of data online in 2015. To put that in perspective, that’s roughly 100 billion gigabytes, or 6.25 billion of the lowest storage iPhone 6. A study from Shareaholic said that in 2013, over 20 percent of internet traffic came from Facebook and Pinterest alone. Our private lives are far less so than ever, and the dangers we face come from criminals, opportunists, and even our fellow man. These tech gurus are focusing on protecting everything they can in our digital lives.

… In one of the papers being presented at CHI, Microsoft researchers asked 34 people to monitor their blood pressure five times a day, in order to better understand whether it would be useful to have continuous blood pressure monitoring in a gadget such as Microsoft Band.

They found that while people were interested in their blood pressure, they didn’t necessarily understand what the data meant.

The National Football League announced a multi-year agreement with Sportradar US. Beginning in the 2015, NFL season, Sportradar US will become the NFL’s exclusive distributor of comprehensive statistics to digital outlets. It will also become the sole provider for the NFL’s own Next Gen Stats, as well as the Game Statistics and Information System (GSIS) which is comprised of official real-time scores, player statistics and play-by-play data.

Google appears to be an indomitable force. But, with today’s release from the US military’s research arm of its Memex search technologies and Europe’s competition investigation into the Mountain View giant, it might be a propitious time for tech-minded entrepreneurs to start building a Google killer.

DARPA’s Memex search technologies have garnered much interest due to their initial mainstream application: to uncover human trafficking operations taking place on the “dark web”, the catch-all term for the various internet networks the majority of people never use, such as Tor, Freenet and I2P. And a significant number of law enforcement agencies have inquired about using the technology. But Memex promises to be disruptive across both criminal and business worlds.

… Bio Inspired Technologies of Boise, Idaho, … is building a sense-and-avoid system using a memristor, a resistor with a memory. Like the synapse in a biological brain, the memristor changes when impulses pass through it. Crucially, it is able to remember the impulse after it has stopped.

This capability forms the basis of a learning system that mimics neurons and the connections between them. A chip-sized neural system linked to the drone’s existing camera can be trained to recognise aircraft and other hazards at long range. Bio Inspired’s drone should be ready for its first flight later this year.

The Health System Measurement Project tracks government data on critical U.S. health system indicators. The website presents national trend data as well as detailed views broken out by population characteristics such as age, sex, income level, and insurance coverage status.

IBM has established a research group in San Jose, Calif., to work on what could be the next big thing in artificial intelligence software. This news came on April 8 after 100 people were assigned to the project, according to the MIT Technology Review. The IBM team will work on algorithms created by Jeff Hawkins, founder of Numenta.

Hawkins spent time creating a theory to explain the inner workings of the human brain, and then applied the concepts to a software blueprint. His algorithms operate in a network, aimed at recreating the behavior of repeating circuits of about 100 neurons in the brain. These neurons can be found on the outer layer of the brain in the neocortex.

We’ve heard about the Three Laws of Data Capital from Paul Sonderegger at Oracle: data comes from activity, data tends to make more data, and platforms tend to win. Advanced analytics enables enterprises to take full advantage of the data their activity produces, ranging from IoT sensors and PoS transactions to social media and image/video. Traditional BI tools produce summary data from data – producing more data, but traditional BI tools provide a view of the past – what did happen. Advanced analytics also produces more data from data, but this data is transformative, generating previously unknown insights and providing a view of future behavior or outcomes – what will likely happen. Oracle provides a platform for advanced analytics today through Oracle Advanced Analytics on Oracle Database, and Oracle R Advanced Analytics for Hadoop on Big Data Appliance to support investing data.

Psychology researchers have recently found themselves engaged in a bout of statistical soul-searching. In apparently the first such move ever for a scientific journal the editors of Basic and Applied Social Psychology announced in a February editorial that researchers who submit studies for publication would not be allowed to use a common suite of statistical methods, including a controversial measure called the p-value.

These methods, referred to as null hypothesis significance testing, or NHST, are deeply embedded into the modern scientific research process, and some researchers have been left wondering where to turn. “The p-value is the most widely known statistic,” says biostatistician Jeff Leek of Johns Hopkins University. Leek has estimated that the p-value has been used at least three million scientific papers. Significance testing is so popular that, as the journal editorial itself acknowledges, there are no widely accepted alternative ways to quantify the uncertainty in research results—and uncertainty is crucial for estimating how well a study’s results generalize to the broader population.

I spend most of my time worrying about how to make deep learning with neural networks faster and more power efficient. In practice that means focusing on a function called GEMM. It’s part of the BLAS (Basic Linear Algebra Subprograms) library that was first created in 1979, and until I started trying to optimize neural networks I’d never heard of it. To explain why it’s so important, here’s a diagram from my friend Yangqing Jia’s thesis: