IBM Zurich researchers have developed a generic artificial-intelligence preprocessing building block for accelerating Big Data machine learning algorithms by at least 10 times over existing methods. The approach, which IBM presented Monday (Dec. 4) at the Neural Information Processing Systems conference (NIPS 2017) in Long Beach, Calif., uses mathematical duality to cherry-pick the items in a Big Data stream that will make a difference, ignoring the rest.
In this episode Byron [Reese] and Pedro Domingos talk about the master algorithm, machine creativity, and the creation of new jobs in the wake of the AI revolution. [audio, 53:50]
Carnegie Mellon University, Dietrich College of Humanities and Social Sciences
from
Carnegie Mellon University engineers and cognitive neuroscientists have demonstrated that a new high-density EEG can capture the brain’s neural activity at a higher spatial resolution than ever before.
This next generation brain-interface technology is the first non-invasive, high-resolution system of its kind, providing higher density and coverage than any existing system. It has the potential to revolutionize future clinical and neuroscience research as well as brain-computer interfaces.
To test the custom-modified EEG, the research team had 16 participants view pattern-reversing black and white checkerboards while wearing the new “super-Nyquist density” EEG. They compared the results from all electrodes to results when using only a subset of the electrodes, which is an accepted standard for EEG density. Published in Scientific Reports, the results showed that the new “super-Nyquist” EEG captured more information from the visual cortex than any of the four standard “Nyquist density” versions tested.
“The U.S. National Institutes of Health has noticed something: More of the research it’s funding is ending up in questionable journals. Recently, the agency issued a statement highlighting some qualities of these journals — aggressively soliciting submissions, failing to provide clear information about pricing — and urging researchers to avoid them. The NIH’s goal: to “help protect the credibility of papers arising from its research investment.” We asked the NIH for more information about the guide notice; a representative returned responses, asking that we attribute them to the NIH Office of Extramural Research.”
The Belfer Center at the Harvard Kennedy School of Government has published a nonpartisan Cybersecurity Campaign Playbook, as part of its Defending Digital Democracy project (previously, “designed to give you simple, actionable information that will make your campaign’s information more secure from adversaries trying to attack your organization—and our democracy.”
There’s a new app available for healthcare patients to make some money off their medical data.
Falls Church, Va.-based healthcare IT company Health Wizz has created a patient-data-aggregation platform that allows patients to trade and sell their data to pharmaceutical companies, researchers and other organizations.
The platform, which was previously available in beta and was relaunched Thursday, runs on a mobile app through which patients can aggregate their health records.
Hartford InsurTech Hub, the accelerator launched by Startupbootcamp in September, has announced which 11 startups will participate in the incubator’s first program, beginning Jan. 8. The insurtechs selected for the three-month accelerator were chosen from a list of 1,000 applicants, according to the company. Each will receive initial $25,000 cash grants and access to founding industry partners—Cigna, Travelers, USAA and The Hartford—as potential investors. This year’s cohort focuses heavily on AI, machine learning and big data to improve customer experience and retention.
Today’s infographic from Mergalim shows how the power of predictive data analytics is growing, and using tools like AI and Bayesian inference to anticipate the outcome of events before they happen is more feasible and applicable than ever.
Almost 15 years after scientists first sequenced the human genome, making sense of the enormous amount of data that encodes human life remains a formidable challenge. But it is also precisely the sort of problem that machine learning excels at.
On Monday, Google released a tool called DeepVariant that uses the latest AI techniques to build a more accurate picture of a person’s genome from sequencing data.
DeepVariant helps turn high-throughput sequencing readouts into a picture of a full genome. It automatically identifies small insertion and deletion mutations and single-base-pair mutations in sequencing data.
This week legal AI company, Kira Systems, announced that it is being used by White Shoe, New York law firm, Davis Polk & Wardwell for transactional doc review. A day earlier, another global M&A powerhouse, Latham & Watkins, also told the world they were using AI for doc review, also in this case, Kira.
But, this isn’t a Kira alone story. Far from it. UK-based Luminance has gained an extra $10m in funding and now opened in the US. Eigen Technologies is already so well established it’s hiring the likes of former Slaughter and May partners to be internal legal advisers. eBrevia and many others are working with law firms all around the world, from San Francisco to Singapore. And Seal goes from strength to strength providing AI doc review and analysis services to the world’s largest corporates.
We had two papers accepted to the conference this year, and contributed to two others. The researchers involved are among the 120+ people on the NVIDIA Research team focused on pushing the boundaries of technology in machine learning, computer vision, self-driving cars, robotics, graphics, computer architecture, programming system, and other areas.
Although working across a variety of fields, they share common goals: advance the science of AI, develop new tools and technologies that will lead to more breakthroughs, and apply AI to modern day grand challenges like autonomous vehicles and healthcare.
A team from Imperial College London has made what is called a significant step forward by reducing the distance over which light can interact by 10,000 fold. This, says the team, means that optical processing could be integrated into chips.
Dr Michael Nielsen, from the Department of Physics at Imperial, said: “This research has ticked one of the boxes needed for optical computing. Because light does not easily interact with itself, information sent using light must be converted into an electronic signal, then back into light. Our technology allows processing to be achieved purely with light.”
Reuters outlines how it has almost entirely automated the identification of breaking news stories. Xiaomo Liu and pals at Reuters Research and Development and Alibaba say the new system performs well. Indeed, it has the potential to revolutionize the news business. But it also raises concerns about how such a system could be gamed by malicious actors.
The new system is called Reuters Tracer. It uses Twitter as a kind of global sensor that records news events as they are happening. The system then uses various kinds of data mining and machine learning to pick out the most relevant events, determine their topic, rank their priority, and write a headline and a summary. The news is then distributed around the company’s global news wire.
Columbia Journalism Review, Duncan Watts and David Rothschild
from
We agree that fake news and misinformation are real problems that deserve serious attention. We also agree that social media and other online technologies have contributed to deep-seated problems in democratic discourse such as increasing polarization and erosion of support for traditional sources of authority. Nonetheless, we believe that the volume of reporting around fake news, and the role of tech companies in disseminating those falsehoods, is both disproportionate to its likely influence in the outcome of the election and diverts attention from the culpability of the mainstream media itself.
To begin with, the breathlessly repeated numbers on fake news are not as large as they have been made to seem when compared to the volume of information to which online users are exposed.
New Orleans, LA The Seventh Joint Conference on Lexical and Computational Semantics will be June 5-6 2018, Co-located with NAACL 2018. Deadline for paper submissions is March 2, 2018.
When I started to implement SoS Notebook as a Jupyter kernel for a Python3-based workflow engine called SoS, it became obvious to me that I needed to expand the Jupyter frontend to provide a more comprehensive user interface for interactive multi-language data analysis. Although SoS Notebook is based on a workflow engine, you can use SoS Notebook as a Python 3 kernel with multi-language support because SoS is implemented as an extension to Python 3.6. The SoS workflow engine greatly expands the usability of SoS Notebook but it is a complete topic by itself and deserves a separate post.
The rest of this post will explain key features of SoS Notebook in details.
Scientists only trust an empirical finding when (at minimum) it can be consistently replicated by independent researchers in independent samples. Curate Science is a crowdsourced platform to track, organize, and interpret replications of published findings in the social sciences.
