We investigate the choice of tuning parameters for a Bayesian multi-level group lasso model developed for the joint analysis of neuroimaging and genetic data. The regression model we consider relates multivariate phenotypes consisting of brain summary measures (volumetric and cortical thickness values) to single nucleotide polymorphism (SNPs) data and imposes penalization at two nested levels, the first corresponding to genes and the second corresponding to SNPs. Associated with each level in the penalty is a tuning parameter which corresponds to a hyperparameter in the hierarchical Bayesian formulation. Following previous work on Bayesian lassos we consider the estimation of tuning parameters through either hierarchical Bayes based on hyperpriors and Gibbs sampling or through empirical Bayes based on maximizing the marginal likelihood using a Monte Carlo EM algorithm. For the specific model under consideration we find that these approaches can lead to severe overshrinkage of the regression parameter estimates in the high-dimensional setting or when the genetic effects are weak. We demonstrate these problems through simulation examples and study an approximation to the marginal likelihood which sheds light on the cause of this problem. We then suggest an alternative approach based on the widely applicable information criterion (WAIC), an asymptotic approximation to leave-one-out cross-validation that can be computed conveniently within an MCMC framework.

This article outlines a method for automatically generating models of dynamic decision-making that both have strong predictive power and are interpretable in human terms. This is useful for designing empirically grounded agent-based simulations and for gaining direct insight into observed dynamic processes. We use an efficient model representation and a genetic algorithm-based estimation process to generate simple approximations that explain most of the structure of complex stochastic processes. This method, implemented in C++ and R, scales well to large data sets. We apply our methods to empirical data from human subjects game experiments and international relations. We also demonstrate the method's ability to recover known data-generating processes by simulating data with agent-based models and correctly deriving the underlying decision models for multiple agent models and degrees of stochasticity.

A report of drone sightings from the Federal Aviation Administration (FAA) shows that despite a new registration scheme, near misses between unmanned and piloted aircraft in American are on the rise. Sightings by pilots and airport officials have steadily increased from less than one a day in 2014, to over 3.5 between August 2015 and January this year, many of them from commercial passenger aircraft. In the most serious incident, the pilot of an American Airlines jet last September had to swerve to avoid a drone. On September 13, flight 475 took off from Atlanta, Georgia en route to Charlotte, North Carolina. It was climbing to 3,500 ft when the pilot of the Airbus had to take evasive action to avoid a collision with an unidentified unmanned aerial system (UAS) or drone.

A drone has successfully delivered a package to a residential location in a Nevada town in what its maker and the state's governor said on Friday was the first fully autonomous urban drone delivery in the US. Matt Sweeney, chief executive of drone-maker Flirtey, said the six-rotor drone flew about a half-mile along a programmed delivery route on 10 March, then lowered the package outside a vacant residence in Hawthorne. The route was established using GPS. A pilot and visual observers were on standby during the flight but were not needed, Sweeney said. He said the package included bottled water, food and a first-aid kit.

Every weekend we select a handful of in-depth articles we think are worth a bit of your valuable time, either because they peel back the layers on a compelling business story, or somehow make us look at business in a different light. AI may undermine big-company advantages. Machine learning – software that can improve itself without human intervention – may mean trouble for big companies that depend on their heft to outmaneuver smaller upstarts, writes Howard Yu for the Harvard Business Review. And for a sneak preview of where the world is headed, one need not look further than the success story of AlphaGo, an artificial intelligence that beat a champion of the ancient game of Go, something that was previously thought to be impossible. "It is easy to imagine a world where self-taught algorithms will play a much bigger role in coordinating economic transactions; AlphaGo simply shows us what is possible in the near future. With instantaneous adjustment, automatic optimization, and continuous improvement all quietly managed by unsupervised algorithms, the redundancy of production facilities and wastage in the supply chain should become headaches of the past."

Lately there seems to be a surge in the term machine learning. Much like big data a few years ago, machine learning is the new buzzword -- and the two terms actually go hand in hand. With increasing volumes of data now stored in distributed environments such as Hadoop, it's possible to quickly produce models that can analyze bigger, more complex data, and deliver faster and more accurate results – two critical elements in the battle against fraud. The more time it takes to discover an instance of fraud, the more the victim organization loses. Association of Certified Fraud Examiners (ACFE) estimates fraud costs organizations 5 percent of annual revenues worldwide.

Technology is the future of retail. Digital data, machine learning, cloud-powered analytics and the Internet of Things (IoT) will separate the wheat from the chaff in tomorrow's retail industry. A recent Sector Insights government report (PDF) said that retailers will need to embrace the latest technology trends, such as big data, and have the skills to work with digital systems if they are to be successful in the future. The retail industry is on the whole a voracious adopter of modern data-centric technology, aping the manufacturing world by using big data analytics to streamline supply chains, and using smartphone apps and wireless beacons to harvest customer data to deliver better service. However, Robert Hetu, retail research director at analyst house Gartner, noted that, despite having the technology to collect and access large amounts of digital data, retailers fail to put it to effective use.

Microsoft recently unveiled Tay, an artificial intelligent chat bot developed by Microsoft's Technology and Research and Bing teams to experiment with and conduct research on conversational understanding. The company stated that the more you chat with Tay, "the smarter it gets, learning to engage people through casual and playful conversation." Microsoft launched a verified Twitter account for "Tay" – billed as its "AI fam from the internet that's got zero chill". However, pretty soon after Tay launched, people starting tweeting the bot with all sorts of misogynistic, racist, and Donald Trumpist remarks. And as Tay was being essentially a robot parrot with an internet connection, started repeating these sentiments back to users.

When dealing with machine learning problems, sometimes one has to face a huge dataset with hundreds or thousands of features. Machine learning relies on these data to build models for prediction, more information the features contain, more easier to train a good model. However, these variables also contain noise, and most of them might be anonymous or formatted by some kind of hash process due to privacy issue or confidential reasons. So it's hard to figure out the physical meaning and explain the correlation between these variables. Even we could know all variables' meanings, it's still difficult to determine which are more essential than others.

Over July 4th weekend in 1981, several hundred game nerds gathered at a banquet hall in San Mateo, California. Personal computing was still in its infancy, and the tournament was decidedly low-tech. Each match played out on a rectangular table filled with paper game pieces, and a March Madness-style tournament bracket hung on the wall. The game was called Traveller Trillion Credit Squadron, a role-playing pastime of baroque complexity. Contestants did battle using vast fleets of imaginary warships, each player guided by an equally imaginary trillion-dollar budget and a set of rules that spanned several printed volumes. If they won, they advanced to the next round of war games--until only one fleet remained. Doug Lenat, then a 29-year-old computer science professor at nearby Stanford University, was among the players. But he didn't compete alone. He entered the tournament alongside Eurisko, the artificially intelligent system he built as part of his academic research. Eurisko ran on dozens of machines inside Xerox PARC--the computer research lab just down the road from Stanford that gave rise to the graphical user interface, the laser printer, and so many other technologies that would come to define the future of computing. That year, Lenat taught Eurisko to play Traveller. Doug Lenat says his common-sense engine is a new dawn for AI. The rest of the tech world doesn't really agree with him. Lenat fed the massive Traveller rulebook into the system and asked it to find the best way of winning.