NYU Data Science newsletter – February 26, 2016

GROUP CURATION: N/A

… Brick-and-mortar retailers can replicate ideas that enterprise retailers are doing in their microcosm with the right software, and if they are linked to a cloud POS solution. These platforms enable the Big Data that’s required where APIs can perform specialized functions that can generate ROI physical retail.

Without better insights and analytics, retail owners often don’t have the resources to do marketing that’s effective, especially if the store scales quickly. That’s why SaaS add-ons that plug into Cloud POS data, are so optimal.

In the wake of the credit markets implosion in 2008, the Global Systemically Important Banks (G-SIB) were faced with a multitude of challenges. Challenges around risk management, reputation, financial stability – and not least regulation. The new mantra from regulators was to ensure that the G-SIB could endure extreme volatility and prolonged periods of economic stress.

One such set of regulations are the Basel accords. With Basel, standardized rules-based approaches are available as well as the option to use data-driven modeling approaches.

The second part is key.

By adopting a data-driven approach, G-SIB institutions can provide a significant reduction in capital requirements, more visibility into ongoing risk levels and more options for improvement such as greater collateralization and optimal product mixes. With thousands of people doing model development and hundreds of people validating those models across various stages, this is a laborious undertaking by any measure with fragmented teams spread across different locations utilizing disparate analytics approaches.

We recently notes 8 Ways Machine Learning Will Improve Education. But even closer to home, this post was inspired by a text from my iPhone which had inferred by day and time where I was headed and, after checking the traffic, let me know I’d be home in five minutes. Half awesome, half spooky, machine learning, according to Domingos, is the new infrastructure for everything.

Following are 10 ways data science, and specifically machine learning, are making your life better.

1. Machine learning is modeling cancer to find a cure.

Growing urbanization presents the United States with an opportunity to showcase its innovation strength, grow its exports, and help to improve citizens’ lives – all at once. Seizing this triple opportunity will involve a concerted effort to develop and apply new technologies to enhance the way cities work for the people who live there.

A new report released today by the President’s Council of Advisors on Science and Technology (PCAST), Technology and the Future of Cities, lays out why now is a good time to promote technologies for cities: more (and more diverse) people are living in cities; people are increasingly open to different ways of using space, living, working, and traveling across town; physical infrastructures for transportation, energy, and water are aging; and a wide range of innovations are in reach that can yield better infrastructures and help in the design and operation of city services.

As part of Sam Woolley’s provocateur-in-residence workshop at Data & Society, I was asked to write a provocation regarding automated agents and bots from my perspective as a poet and artist. In response, I offer the following explanation of what, in my view, a bot is. After all, what in this world is more provocative than a definition?

Smart city projects could help municipal governments operate more efficiently and improve quality of life for residents, but data access, privacy and security are still hurdles to overcome.

It was during the financial crisis that Andrew Lo had his epiphany: The way to save health care from ever-rising costs is by bringing in the banks. Specifically, by packaging drug development costs into securities to be bought and sold by Wall Street—the very, um, mortgage-bundling technique that blew up the economy in 2007. “The reason the financial crisis happened is not because securitization didn’t work. It happened because it worked way too well,” says Lo, a professor of financial engineering at MIT. Securitization injected a huge pool of money into mortgages—what if you could inject that pool of money into a worthwhile cause and, ahem, do it responsibly?

So Lo, who has seen his mother and several friends die from cancer, wants to use the techniques of Wall Street to fix healthcare. In a new paper in Science Translational Medicine, he and his coauthors propose creating loans for patients whose insurance policies don’t cover ultra-expensive treatments like the cure for hepatitis C—loans that would be financed by bundling them and selling to Wall Street investors.

Back-office plumbing will take a front seat in LSE-Deutsche Boerse’s potential tie-up.

The business of clearing and settling trades has always looked dull compared with the racier business of buying and selling. But it’s likely to be at the heart of whether regulators approve a tie-up between Deutsche Boerse and London Stock Exchange — and whether the financial system will be riskier or safer as a result.

With his new book, Dark Territory: The Secret History of Cyber War, Kaplan dives into a topic which could end up being just as transformational to national security affairs as the nuclear age was. The book opens fast and builds from there, providing insights from research that even professionals directly involved in cyber operations will not have gleaned. Famed author John le Carre described it this way: “A book that grips, informs, and alarms, finely researched and lucidly related”. I have to agree.

The reason to study the past is to inform today’s decisions, and there is a dire need for this type of research today. The discipline of cyber warfare is still very young and the people charged with leadership in this domain are too frequently left to their own to discover the history of their profession. This book captures all the key elements of foundational cyber conflict knowledge in a way that can help today’s national security strategists operate and plan for the future.

Creating a data set with high spatial resolution for some of the countries that could benefit from better internet connectivity is a large undertaking. Aggregate population counts on the spatial scale of provinces or districts are known from population censuses but alone are insufficient, as these areas vary in geographical size and do not provide insight about population distributions on a granular level.

We solved this challenge by applying techniques from computer vision on DigitalGlobe high-resolution satellite imagery. We identified human-built structures, such as buildings or other infrastructure, and used those locations as a proxy for where people live. We then combined our results with existing census counts and created a population data set with 5-meter resolution for 20 countries. While recognizing structures in aerial imagery is a popular task in computer vision, scaling it to a global level came with additional difficulty.

Facebook has announced it will make highly detailed maps of places where it believes people are living available to the public later this year.

The social network has been using artificial intelligence software to scan satellite imagery and identify human-built structures.

It hopes to use the information to determine where internet-beaming drones would best be deployed.

Over the years, Hilary Mason, founder of Fast Forward Labs, visited theCUBE, from the SiliconANGLE Media team, to share her thoughts on data and how it will impact the enterprise. The segments below provide a glimpse into the mind of a true data scientist. [3 videos; 13:13, 16:17, 16:53]

A team of ISTC researchers is developing a new polystore, BigDAWG, to incorporate a wide variety of engines and data models, such as the relational model with PostgreSQL, an array model with SciDB, a key-value model with Accumulo, and a streaming model with S-Store. A promising alternative approach is to use just-in-time operators on raw data, such as NoDB and ViDa, to virtualize data sources.

To balance the trade-off between functionality and database-transparency, a BigDAWG query specifies Islands of Information that provide a set of operations and data model. The system manages data transformation between models and subsequently engine placement. Decisions about data placement consider required functionality, workload behavior and expected performance.