NYU Data Science newsletter – June 2, 2015

GROUP CURATION: N/A

For a concept that we all understand, ‘simple’ is deceivingly difficult to pin down.

We may ‘know it when we see it’, but there’s more to what makes a product or website feel simple than just gut reaction.

I really like the enthusiastic and motivating way he teaches the lectures. (I just watched the first lesson, so no real material yet.)

Can anyone tell me how deep this course dives into the theory? I plan on doing all the homework also, he posted some. His coursera course is said to be very practical, and I want to really understand the theory. Will this lecture get me started off with understanding the fundamental concepts very well? For example, would I be able to understand what the different chapters of Bishop’s book are about and be able, if I wanted to, to just go to the chapter I want to learn more about and read that?

pyLDAvis is designed to help users interpret the topics in a topic model that has been fit to a corpus of text data. The package extracts information from a fitted LDA topic model to inform an interactive web-based visualization.

Smartphones are pretty good at understanding what users say to them, but they can’t handle conversations. The shift from one speaking voice to another confuses virtually all current speech-recognition software. But researchers at IBM IBM -0.06% have succeeded in building an algorithm that doesn’t get tripped up so easily. The new software, said Michael Picheny, the leader of IBM’s speech team, accurately recognizes conversations spoken by two alternating voices.

The big players in financial services have long struggled to keep pace with technology. They’ve been slow to deploy the latest in computers and software to analyze data, pick investments, and engage customers.

That has created big opportunities for a slew of local entrepreneurs.

“The industry is just crying out for new ways of doing things,” said Adam Broun, chief operating officer of Kensho Technologies Inc., a Cambridge software company launched two years ago to speed the process of modeling the effect of world events on the prices of stocks, bonds, and other assets.

The Alfred P. Sloan Foundation is supporting a project by Data & Society to better understand the needs of computer science researchers as they navigate emerging issues of privacy, ethics, and equitable access to data at different phases of the research process.

New and complex data sets raise challenging ethical questions about risk to individuals that are not sufficiently covered by computer science training, ethics codes, or Institutional Review Boards. The use of publicly available, corporate, and government datasets may reveal human practices, behaviors, and interactions in unintended ways, creating the need for new kinds of ethical support.

We’re delighted to announce that Sergey Levine, who works at the intersection of robotics, machine learning, graphics, and animation, will join the UW CSE faculty.

Sergey pioneered the recent trend in using deep learning to create neural network controllers for animated characters and robots. His learning techniques enable robots to solve control tasks that have been elusive using traditional approaches. This past week he won the Best Robotic Manipulation Paper Award at ICRA, the IEEE flagship robotics conference, for his work on learning controllers for complex manipulation tasks.

In this exclusive interview, [Ben] Porterfield explains how all founders can nail down an analytics infrastructure from the very beginning — and shares his wisdom on where to store data, the best tools to use, common mistakes to dodge, and what you should measure to start making the right moves today.