NYU Data Science newsletter – July 6, 2016

GROUP CURATION: N/A

Amazon is always on the lookout for new robotic technologies to improve efficiency in its warehouses, and this year deep learning appears to be leading the way.

That’s according to the results of the second annual Amazon Picking Challenge, which has been won by a joint team from the TU Delft Robotics Institute of the Netherlands and the company Delft Robotics.

Machine learning techniques based on neural networks are achieving remarkable results in a wide variety of domains. Often, the training of models requires large, representative datasets, which may be crowdsourced and contain sensitive information. The models should not expose private information in these datasets. Addressing this goal, we develop new algorithmic techniques for learning and a refined analysis of privacy costs within the framework of differential privacy. Our implementation and experiments demonstrate that we can train deep neural networks with non-convex objectives, under a modest privacy budget, and at a manageable cost in software complexity, training efficiency, and model quality.

National Privacy Research Strategy calls for research in science and engineering that will enable the U.S. to benefit from innovative data use while protecting privacy.

There seems to be very little overlap currently between the worlds of infosec and machine learning. If a data scientist attended Black Hat and a network security expert went to NIPS, they would be equally at a loss.

This is unfortunate because infosec can definitely benefit from a probabilistic approach but a significant amount of domain expertise is required in order to apply ML methods.

Machine learning practitioners face a few challenges for doing work in this domain including understanding the datasets, how to do feature engineering (in a generalizable way) and creation of labels.

Raj Acharya sits in the warm morning sun just a few yards from the Rose Well House on the Old Crescent at Indiana University, one of IU’s earliest features meeting one of the university’s newest arrivals. He gazes at the iconic statue of former president Herman B Wells, his voice quiet but confident, the voice of a man who is ready to tackle his next challenge.

“This is very exciting,” the new Dean of the School of Informatics and Computing says as a slight smile begins to spread across his face. “It’s history repeating.”

Acharya’s path to Bloomington began more than 8,000 miles away in Bangalore, India

Google’s DeepMind is teaming up with NHS-funded Moorfields Eye Hospital to research whether machine learning can help the fight against blindness. … Machine learning processes will be applied to around a million eye scans to help search for early symptoms of sight loss.

A school-based program with quadrivalent human papillomavirus (HPV) vaccination was implemented in Alberta in 2008. We assessed the impact of this program on Pap test cytology results using databases of province-wide vaccination and cervical cancer screening.

A warehouse in China barely needs any human workers to function. BBC Click went to Shanghai to see how it works.

Computational approaches for creating machines at scale: An interview with Daniela Rus.

Clara Labs is establishing a class of coworker that combines the best of human and machine talents. Like a machine, Clara always remembers preferences. Like a good assistant, Clara appreciates the nuance underlying each of your messages and helps build rich relationships as a team member. To achieve this automated-yet-socially-adept experience, person and machine work and learn in tandem to complete tasks in a positive feedback loop that we think of as cooperative intelligence.

Our platform enables Clara remote assistants (CRAs) to be super-human: they can handle complex requests from hundreds of customers with little a priori background on each customer or conversation.

AI is primarily concerned with the creation of intelligent agents that use methods and techniques such as machine learning, optimization, language processing, logic and search to automate otherwise manual tasks. Most of the AI startups here are pitching at business use (usually sales and marketing), but a few are also developing consumer products (e.g. Anki). Applications in healthcare are also becoming common.

Washington, DC OpenCon is the conference and community for students and early career professionals interested in advancing Open Access, Open Education and Open Data. OpenCon 2016 will be held on November 12-14 in Washington, DC. Attendance at OpenCon 2016 is by application only.

Deadline for applications is Monday, July 11.

Selecting effective colors for bar plots, pie charts, and heat maps is made more difficult by the fact that the way we select color in software does not reflect how we perceive the color.

There are many examples of poor color combinations in published figures. For example, if categories are encoded with a combination of bright and dark colors, the bright colors will dominate the reader’s attention. On the other hand, if two colors appear similar, the reader will instinctively perceive them as belonging to a group and infer that the underlying variables are related.

This document describes how I set up my developer environment on a new MacBook or iMac. We will set up Node (JavaScript), Python, and Ruby environments, mainly for JavaScript and Python development. Even if you don’t program in all three, it is good to have them as many command-line tools use one of them. As you read and follow these steps, feel free to send me any feedback or comments you may have.

Extended geographic projections for D3.