Artificial intelligence has replaced big data this year as the most talked about new set of technologies. As with big data five years ago--behind the hype, the confusion generated by an ill-defined term, and the record funding by VC--we are starting to see emerging investments and practical applications where it matters most--in enterprises. A new report from Narrative Science, based on a survey of 235 business executives conducted by the National Business Research Institute (NBRI), sheds light on the state-of-AI in enterprises today and in the future: 38% of enterprises are already using AI technologies and 62% will use AI technologies by 2018. Keep in mind that "AI technologies" is a broad term that includes machine and deep learning, recommendation engines, predictive and prescriptive analytics, automated written reporting and communications, and voice recognition and response. Big data has spawned the current interest and increased investment in artificial intelligence.

Researchers used a combination of three imaging techniques and a machine learning system to create this new map of the brain, which includes 180 distinct regions in each hemisphere of the cerebral cortex (the brain's outermost region). The different colors relate to how connected an area is to a specific sensory input: Red is hearing, green is touch, and blue is vision. Mixed colors represent areas where two senses overlap. A group of researchers have developed a new map of the cerebral cortex of the human brain, revealing 100 new distinct regions in each hemisphere. Representing the most detailed map of the brain yet, it's an achievement of a longstanding goal, and researchers say it will provide a crucial tool to understand how differences in even extremely small brain regions relate to behavior and disease.

The 4U4 card design of Inspur NF5568M4 is applicable to present electric power and heat dissipation designs of the data center, and is scalable to multi-machine and multi-card CPU computing clusters via the open-source Inspur Caffe-MPI becoming the mainstream CPU server used presently in the internet industry. Currently, Inspur's deep learning solution is being applied at Tencent, Baidu, Alibaba, Qihoo, iFLYTEK and JD and is supporting the "super brains" of various types of intelligentized services. As the neural network model grows in complexity, the computing performance necessary for the model training increases dramatically. The cluster-edition Caffe-MPI computing frame launched by Inspur achieves parallel computing of GPU server. It adopts high-performance mature MPI technology in computing -- carrying out parallel data optimization to the Caffe edition -- with the ability to organize multiple NF5568M4 into CPU parallel computing clusters via IB network.

Chinese technology company Baidu today announced the launch of a few new services within its Baidu Open Cloud public cloud infrastructure portfolio. Baidu TianSuan (Smart Big Data) lets customers "collect, store, process and analyze big data," Baidu said in a statement. Baidu TianXiang (Smart Multimedia Cloud) includes face recognition and live video streaming, while Baidu TianGong (Intelligent IoT Service) is a full-stack platform for integrating cloud applications with internet-connected devices. The additions bring Baidu more in line with the world's leading cloud infrastructure providers, including Amazon Web Services, Microsoft Azure, and Google Cloud Platform. Amazon and Microsoft have both introduced Internet of Things (IoT) services.

WASHINGTON – Neuroscientists acting as cartographers of the human mind have devised the most comprehensive map ever made of the cerebral cortex, the part of the brain responsible for higher cognitive functions such as abstract thought, language and memory. Using MRI images from the brains of 210 people, the researchers said on Wednesday they were able to pinpoint 180 distinct areas in the cerebral cortex, the brain's thin, wrinkly outermost layer made of so-called gray matter. These areas were present in both the left and right hemispheres of the cerebral cortex. More than half, 97 of them, were previously unknown. The researchers nailed down the specific function of some of the areas, but said they were only scratching the surface on understanding what all of the areas did.

The post highlights points raised by Yoav in his official response to the ASA statement (available as on page 4 in the ASA supplemental tab), as well as offers a list of relevant R resources. It is just as well relevant to the use of most other statistical methods: context matters, no single statistical measure suffices, specific thresholds should be avoided and reporting should not be done selectively. The latter problem is discussed mainly in relation to omitted inferences. We argue that the selective reporting of inferences problem is serious enough a problem in our current industrialized science even when no omission takes place. Many R tools are available to address it, but they are mainly used in very large problems and are grossly underused in areas where lack of replicability hits hard.

The release continues, "PhishAlarm Analyzer scans reported emails and examines them based on the attributes of the email and linguistic characteristics of the text. The emails are then prioritized, and an HTML threat report on the suspicious email is delivered to the security and incident response teams. The research report saves time for the incident response team by performing much of the research in advance so that they respond more quickly to the reported threats. By using various email threat feeds coupled with machine learning, PhishAlarm Analyzer constantly improves as it learns new patterns of email threats. PhishAlarm Analyzer is built to scan emails quickly and prioritize the threat, including identifying zero-hour phishing attacks in real time. By quickly detecting and ranking the most dangerous threats, PhishAlarm Analyzer allows incident response teams to remediate quickly and efficiently."

This paper is concerned with the numerical solution of model-based, Bayesian inverse problems. We are particularly interested in cases where the cost of each likelihood evaluation (forward-model call) is expensive and the number of un- known (latent) variables is high. This is the setting in many problems in com- putational physics where forward models with nonlinear PDEs are used and the parameters to be calibrated involve spatio-temporarily varying coefficients, which upon discretization give rise to a high-dimensional vector of unknowns. One of the consequences of the well-documented ill-posedness of inverse prob- lems is the possibility of multiple solutions. While such information is contained in the posterior density in Bayesian formulations, the discovery of a single mode, let alone multiple, is a formidable task. The goal of the present paper is two- fold. On one hand, we propose approximate, adaptive inference strategies using mixture densities to capture multi-modal posteriors, and on the other, to ex- tend our work in [1] with regards to effective dimensionality reduction techniques that reveal low-dimensional subspaces where the posterior variance is mostly concentrated. We validate the model proposed by employing Importance Sam- pling which confirms that the bias introduced is small and can be efficiently corrected if the analyst wishes to do so. We demonstrate the performance of the proposed strategy in nonlinear elastography where the identification of the mechanical properties of biological materials can inform non-invasive, medical di- agnosis. The discovery of multiple modes (solutions) in such problems is critical in achieving the diagnostic objectives.

This paper characterizes hierarchical clustering methods that abide by two previously introduced axioms -- thus, denominated admissible methods -- and proposes tractable algorithms for their implementation. We leverage the fact that, for asymmetric networks, every admissible method must be contained between reciprocal and nonreciprocal clustering, and describe three families of intermediate methods. Grafting methods exchange branches between dendrograms generated by different admissible methods. The convex combination family combines admissible methods through a convex operation in the space of dendrograms, and thirdly, the semi-reciprocal family clusters nodes that are related by strong cyclic influences in the network. Algorithms for the computation of hierarchical clusters generated by reciprocal and nonreciprocal clustering as well as the grafting, convex combination, and semi-reciprocal families are derived using matrix operations in a dioid algebra. Finally, the introduced clustering methods and algorithms are exemplified through their application to a network describing the interrelation between sectors of the United States (U.S.) economy.

In cargo logistics, a key performance measure is transport risk, defined as the deviation of the actual arrival time from the planned arrival time. Neither earliness nor tardiness is desirable for customer and freight forwarders. In this paper, we investigate ways to assess and forecast transport risks using a half-year of air cargo data, provided by a leading forwarder on 1336 routes served by 20 airlines. Interestingly, our preliminary data analysis shows a strong multimodal feature in the transport risks, driven by unobserved events, such as cargo missing flights. To accommodate this feature, we introduce a Bayesian nonparametric model -- the probit stick-breaking process (PSBP) mixture model -- for flexible estimation of the conditional (i.e., state-dependent) density function of transport risk. We demonstrate that using simpler methods, such as OLS linear regression, can lead to misleading inferences. Our model provides a tool for the forwarder to offer customized price and service quotes. It can also generate baseline airline performance to enable fair supplier evaluation. Furthermore, the method allows us to separate recurrent risks from disruption risks. This is important, because hedging strategies for these two kinds of risks are often drastically different.