For the first time, engineers at Caltech have developed an empirical statewide model of the California reservoir network. The model was built from data gathered over a 13-year period that includes the latest drought, allowing researchers to make observations about how 55 of the state’s major reservoirs respond to a variety of external conditions as a unified system.
Company Data Science News
As readers of the newsletter may have been able to predict, there is now more VC funding available for chip start-ups in the AI space. Look out Nvidia and Intel, Graphcore, Mythic, Wave Computing, Cerebras, and Cambricon are coming for your market.
Google can now use facial recognition on its phones to tell if someone is peaking at your phone over your shoulder. Why do I get the feeling this is going to make reading one’s phone on a crowded subway an experience filled with vomit rainbows?
Amazon is a dominant player in the cloud computing business where it holds ~44% of the market share. They are proud of some of their flashiest customers, like the National Football League.
Waymo has clocked 4 million self-driven miles on public roads. No deaths! I’m guessing that’s better than the odds of going 4 million miles in a car driven by a series of humans.
Google is testing technology that would bypass the data entry pain point in electronic medical recording by allowing conversations between doctors and patients to be instantly transcribed into the patient’s record. Sounds efficient, good for science, and creepy as is the case with many new technologies. Ask yourself, data scientists, what could go wrong here?
Optum Ventures, the health services and analytics arm of UnitedHealthcare, is offering a new $250M fund to support start-ups that are improving the healthcare system through AI and data science.
Diane Bryant is leaving Intelto become the chief operating officer for the Google Cloud group at Alphabet.
GE Healthcare will be partnering with Nvidia and with Intel. GE Healthcare is hedging its bets by connecting with two of the main AI chipmaking technologies as it moves into the very hot AI medical imaging space. It looks like the partnership with Intel has less to do with chips and more to do with Intel’s cloud, but as we are learning every week, chips matter!
A team of students and faculty at UW-Seattlewon the inauguralAlexa Prize. The team took home $500,000 in the Amazon-sponsored competition for new conversational artificial intelligence UIs.
Elsewhere at Berkeley, John Tygar is teaching a class called Cyberwar that lets students “forensically examine real cyberwar attacks”. The goal is to teach students how to prevent them, by teasing apart how actual cyber attacks have been orchestrated. Every university should have such a class. Seriously. I’m still pissed about Equifax.
The setup for LIGO—at its two facilities, one in Washington and one in Louisiana—comprises tens of thousands of pieces of equipment, from massive optics to tiny screws. And if anything has so much as a speck of dust on it, it might contaminate LIGO’s optics and diminish the signal of a gravitational wave. Here are some of the things Caltech engineers do to make sure that everything is beyond squeaky clean.
For two tumultuous years, Kathy Partin has led the US Office of Research Integrity (ORI), the government organization tasked with investigating claims of scientific misconduct among federal grant recipients. Retraction Watch reports that as of December 4, Partin will be on leave for three months, having been assigned a temporary position at the Uniformed Services University of the Health Sciences in Bethesda, Maryland. The move marks what could be the end of Partin’s tenure as ORI director.
“They’ve basically told her she’s not going back,” Linda Schutjer, Partin’s former colleague at Colorado State University, tells Retraction Watch. “They’ve told her to take all her stuff out of her office, all of her personal things.” Retraction Watch was unable to confirm Schutjer’s claim with Partin, who declined to comment.
The Nov. 17 award of the 10-year, $50 billion Alliant 2 Large Business contract to 61 companies pits the largest incumbent Alliant Large Business contractors against 28 dynamic new entrants for some of the largest federal information technology orders.
Alliant 2 Large Business (Alliant 2 LB) is a governmentwide acquisition contract (GWAC) managed by the General Services Administration that is used by all federal agencies to buy a wide range of information technology services. Alliant 2 LB’s predecessor — Alliant Large Business — has generated $16.5 billion since fiscal 2009, according to Bloomberg Government data, and is eligible for new orders for another year and a half.
A team of researchers led by Neil L. Kelleher and Philip D. Compton has expanded native mass spectrometry into native proteomics, which they can use in an untargeted discovery mode to analyze endogenous complexes in cells and tissues (Nat. Chem. Biol. 2017, DOI: 10.1038/nchembio.2515). “This technique represents the next step in the evolution of native MS,” Kelleher says.
The U.S. Patent and Trademark Office has revamped its fee schedule for the first time in over four years, raising most patent fees by roughly 10%. PTO says the increases, which will take effect on Jan. 16, 2018, will help cover the office’s operational costs, reduce its backlog, and grow a reserve balance.
There’s an AR game called Internet of Elephants that let’s gamers put the elephants…pretty much anywhere. Like at the mall. Or the office. Or in the elevator.
Robbie Allen is the North Carolina entrepreneur behind the machine learning content generator Automated Insights who left his CEO perch to do a mid-career Ph.D. in Computer Science at the University of North Carolina. He’s leaving his program, boomeranging back to lead a new startup, InfiniaML.
A controversial open-source algorithm designed to flag stats problems in published psych papers claims to have 95% accuracy, but some remain unconvinced. Onwards in the fight to bolster our trust in psychological findings.
“There is this opportunity to build disruptive companies and it is those companies that will ultimately lead to the world looking at Montreal as a centre,” [Andy] Mauro said.
Local companies will also have a bigger benefit to the local economy than large companies from elsewhere that come to Montreal looking for AI talent.
“The bulk of the benefit will not accrue to Montreal,” Mauro said, “if we do not invest in homegrown, headquartered startups here.”
Cartography is the new code. Increasingly, everything from your takeout delivery to your UberPool route is orchestrated not just by engineers but by cartographers. Between 2007 and 2015, the number of grads earning master’s degrees in cartography increased annually by more than 40 percent on average. And as advanced satellites, digital mapping tools, and open-source geographical software progress, the demand for cartographers is projected to grow nearly 30 percent by 2024.
Modern cartographers are as much data analysts as they are map producers. Flagship GIS systems by software companies like Esri have been democratized by an explosion of open-source alternatives like Carto and MapBox. “We are absolutely inundated with volumes of geospatial data,” says Mike Tischler, director of the US Geological Survey’s National Geospatial Program, “but with no means to effectively use it all.”
They seem like odd couples: Aetna, one of the nation’s largest health insurers, is in talks to combine with CVS Health, which manages pharmacy benefits. The Cleveland Clinic, a highly regarded health system, joined forces with an insurance start-up, Oscar Health, to offer individuals a health plan in Ohio.
Aetna also has new partnerships with large health systems that include hospitals and doctors’ groups in Northern California and Virginia.
Former adversaries are banding together, girding against upheaval in a rapidly changing health care environment. They are also bracing for the threats posed by interlopers like Amazon eyeing a foray into the pharmacy business or tech companies offering virtual medical care via a computer or cellphone.
In a series of experiments at Princeton University and the University of California, Los Angeles, students were randomly assigned either laptops or pen and paper for note-taking at a lecture. Those who had used laptops had substantially worse understanding of the lecture, as measured by a standardized test, than those who did not.
The researchers hypothesized that, because students can type faster than they can write, the lecturer’s words flowed right to the students’ typing fingers without stopping in their brains for substantive processing. Students writing by hand had to process and condense the spoken material simply to enable their pens to keep up with the lecture. Indeed, the notes of the laptop users more closely resembled transcripts than lecture summaries. The handwritten versions were more succinct but included the salient issues discussed in the lecture.
Even so, it may seem heavy-handed to ban electronics in the classroom. Most college students are legal adults who can serve in the armed forces, vote and own property. Why shouldn’t they decide themselves whether to use a laptop?
POLITICO, Helena Bottemiller Evich and Catherine Boudreau
from
Nestlé, the world’s largest food company, rocked food circles in late October with the news that it was leaving the industry’s most powerful lobbying group in Washington, the Grocery Manufacturers Association, amid disagreements about how to respond to changing consumer tastes.
The departure of a conglomerate that owns thousands of brands — from Hot Pockets to Deer Park water — was the most visible sign yet that the food industry’s reign in Washington is faltering as some companies scramble to adapt to rapidly evolving consumer demands while others are slower to embrace the trends. Long the attack group for large companies like Kraft and General Mills on legislative and regulatory issues, GMA now has members like Nestlé opposing some of its positions.
NYU Tandon and FutureLabs recently ran an AI Summit (Oct 30-31, 2017) where Cognilytica and AI Today were proud partners. We were able to sit down with Steve Kuyan, a managing Director at FutureLabs to get this thoughts on the Summit, how artificial intelligence is currently being taught in universities, and the knowledge gap that’s taking place around AI. We also talk about the academic brain drain and poke some fun at the concept of the “Chief AI Officer”. [audio, 15:51]
Since 2009, the federal government has disbursed $30 billion in incentive payments to foster the transition to electronic medical records, but no open standard has been in place. 456,000 physicians received some sort of incentive payment. This incentive money inflated the value of contracts, allowing hospital systems to prioritize control of their data over the cost of the system.
Over 80% of the country moved to electronic records, but the data ended up in silos. HIPAA laws and anti-kickback statutes also continue to prevent information exchanges.
During the same period, hospital consolidation continued. Mergers and buyout deal flow never slowed, as hospitals felt they had to get bigger to survive. But this led to a Frankenstein of Electronic Health Record systems on the back end.
In a previous career, Sandra Miarecki flew high above the Earth’s surface. During a 20-year career in the US Air Force that included time as a test pilot, she flew aircraft including the B-52, F-16, MiG-15, helicopters and even the Goodyear Blimp.
She retired from the Air Force in 2007 to pursue a new calling in physics that would set her sights on the depths of the Earth. Now an assistant professor of physics, Miarecki served as the principal researcher in a just-released study that relied on data from a detector encapsulated in ice near the South Pole to determine how high-energy subatomic particles are absorbed as they travel inside the planet.
Seattle, WA January 8-11, 2018, WRF Data Science Studio, 6th floor Physics/Astronomy Tower. This hands-on workshop will cover basic concepts and tools, including program design, version control, data management, and task automation. Software Carpentry will focus on tools for software development (with Python). [$$]
Seattle, WA January 8-11, 2018, WRF Data Science Studio, 6th floor Physics/Astronomy Tower. This hands-on workshop will cover basic concepts and tools, including program design, version control, data management, and task automation. Data Carpentry will focus on tools for data processing and data visualization (with R). [$$]
HFC is a multi-phase program, but your participation can begin now (see below for next steps). If you join HFC as a research participant, you will be randomly assigned to a HFC research system that will compete in a forecasting Tournament to see which system can produce the most accurate forecasts. The formal Tournament will commence in early 2018. In the meantime, starting August 2017 you will be able to hone your forecasting skills by participating in the “HFC Challenge” hosted on the Good Judgment Open (GJO) forecasting site. The HFC Challenge will allow you to explore forecasting procedures, forecast against real-world questions, and check your aptitude for geopolitical & geoeconomic forecasting. While optional, participation in the HFC Challenge is an ideal way to prepare for the Tournament.
In times when fake news and alternative facts circulate in society, spreading scientifically based findings is more important than ever. This makes science communication one of academia’s most vital tasks. But despite the pivotal role scientific communication plays in society, communicating with the general public is not always prioritised among researchers.
This is in part because although scientists tend to be great at doing the research and discovering results, they are not always so great at then communicating these findings to a wider audience. Writing about research for people outside academia requires a wider perspective on science matters, and a completely different writing style. And many scientists may simply not know how to communicate their research to wider society. The low importance placed on this in the past, and the limited (if any) training students get in science education also isn’t helping.
This is shortsighted, particularly as my recent study found, encouraging science students to write about their work for a non academic audience helped them to discover and discuss different ideas within their thesis. And this in turn, helped them to realise the importance and societal impact of their work.
Amazon has launched a new Interactive AWS Cost Explorer API which provides access to underlying Cost Explorer data. The AWS Cost Explorer is a free tool that can be used to view AWS costs and manage AWS Spending. Developers can view AWS spending data for the past thirteen months, and forecast likely expenditures for the next three months.
“My aim here is to test my understanding of Karpathy’s great blog post ‘Hacker’s guide to Neural Networks’ as well as of Python, to get a hang of which I recently perused through Derek Banas’ awesome commented code expositions. As someone steeped in R and classical statistical learning methods for structured data, I’m very new to both Python as well as Neural nets, so it is best not to fall into the easy delusions of competence that stem from being able to follow things while reading about them. Therefore, code.”