We present new techniques for automatically constructing probabilistic programs for data analysis, interpretation, and prediction. These techniques work with probabilistic domain-specific data modeling languages that capture key properties of a broad class of data generating processes, using Bayesian inference to synthesize probabilistic programs in these modeling languages given observed data. We provide a precise formulation of Bayesian synthesis for automatic data modeling that identifies sufficient conditions for the resulting synthesis procedure to be sound. We also derive a general class of synthesis algorithms for domain-specific languages specified by probabilistic context-free grammars and establish the soundness of our approach for these languages. We apply the techniques to automatically synthesize probabilistic programs for time series data and multivariate tabular data. We show how to analyze the structure of the synthesized programs to compute, for key qualitative properties of interest, the probability that the underlying data generating process exhibits each of these properties. Second, we translate probabilistic programs in the domain-specific language into probabilistic programs in Venture, a general-purpose probabilistic programming system. The translated Venture programs are then executed to obtain predictions of new time series data and new multivariate data records. Experimental results show that our techniques can accurately infer qualitative structure in multiple real-world data sets and outperform standard data analysis methods in forecasting and predicting new data.

Program comprehension is a dominant process in software development and maintenance. Experts are considered to comprehend the source code efficiently by directing their gaze, or attention, to important components in it. However, reflecting importance of components is still a remaining issue in gaze behavior analysis for source code comprehension. Here we show a conceptual framework to compare the quantified importance of source code components with gaze behavior of programmers. We use "attention" in attention models (e.g., code2vec) as the importance indices for source code components and evaluate programmers' gaze locations based on the quantified importance. In this report, we introduce the idea of our gaze behavior analysis using the attention map, and the results of a preliminary experiment.

The tests, set to take place at Edwards Air Force Base in Kern County, California, are expected to be conducted on a "small, but representative high-speed surrogate aircraft," Cara Bousie, the service's spokesperson, told Aviation Week. Although Bousie steered clear of offering any additional details regarding the looming tests, she did indicate that the move is part of a two-year campaign for the department to determine just how the technology will perform in a controlled setting. Will Roper, assistant secretary of the Air Force for acquisition, previously revealed in a March interview that aircraft candidates that may be used during the summer trials include the Kratos XQ-58A Valkyrie, Composite Engineering BQM-167 Skeeter and Boeing QF-16. Disclosed to the public just in March, the Skyborg program's objective is to deliver a combat-ready, autonomous, unmanned aerial vehicle prototype by the end of 2023. The aircraft is expected to act as a robotic wingman for service members, using its AI tech to manage combat mission tasks on its own when the need arises.

Even casual music fans can distinguish songs by category without great difficulty, but that's not the case for computers. Most audio-based music classification and tagging systems use categorical supervised learning -- in other words, learning a function that maps songs to genres based on example pairs -- with a fixed set of labels that intrinsically can't handle unseen labels, such as newly added genres. That's why a team of scientists at Naver Corp, an internet content service company headquartered in South Korea, investigated a zero-shot alternative in a paper ("Zero-Shot Learning for Audio-based Music Classification and Tagging") published on the preprint server Arxiv.org. Their AI classification system learns how to recognize songs without any labeled training data by taking into account side information about musical instruments, words in descriptions about songs, and more. The researchers settled on two types of side information at the outset of the study: human-labeled attribute information and general word semantic information.

Megvii, a Chinese AI startup that supplies facial recognition software for the Chinese government's surveillance program, is expanding its technology beyond humans to recognize different faces of pets. As reported by Abacus News, Megvii's new program is trained to recognize dogs by their nose prints -- much like how humans have unique fingerprints. Using the Megvii app, the company says it can register your dog simply by scanning the snout through your phone's camera. Just like how a phone registers your fingerprint for biometric unlocks, the app asks you to take photos of your dog's nose from multiple angles. Megvii says it has an accuracy rate of 95 percent and has reunited 15,000 pets with their owners through the app.

An artificial intelligence (AI)-trained facial recognition system (FRS) has been installed at the Puratchi Thalaivar Dr. MGR Central railway station for detecting known culprits passing through the gates and alerting authorities. "For the first time, we have introduced the CCTV camera device backed by artificial intelligence. In the existing system, we capture the picture and video of any suspect. But we have to manually analyse the footage to detect their movement. The new system will automatically alert us about known culprits," said a senior police officer of the Government Railway Police (GRP).

Preventing unfair treatment of individuals on the basis of race, gender or ethnicity, for example, been a long-standing concern of civilized societies. However, detecting such discrimination resulting from decisions, whether by human decision makers or automated AI systems, can be extremely challenging. This challenge is further exacerbated by the wide adoption of AI systems to automate decisions in many domains -- including policing, consumer finance, higher education and business. "Artificial intelligence systems -- such as those involved in selecting candidates for a job or for admission to a university -- are trained on large amounts of data," said Vasant Honavar, Professor and Edward Frymoyer Chair of Information Sciences and Technology, Penn State. "But if these data are biased, they can affect the recommendations of AI systems."

New update will be the first AI-based cybersecurity solution for the Chrome Operating System, available immediately. With global Chromebook unit shipments slated to nearly double this year to 10.1 million since their release in 2011 (and expected to reach 17 million units in sales by 2023), the Chrome OS is more prevalent in the market than ever before. The new offering makes Deep Instinct the first deep learning based solution with multi-layer protection across endpoints, servers and mobile devices for Windows, macOS, iOS, Android or Chrome OS from the convenience of a single platform. It also protects against a range of attack vectors that have been known to breach Android systems. As the first company to apply end-to-end deep learning to cybersecurity, Deep Instinct provides complete protection against attacks, taking a prediction and prevention first approach, followed by detection and response against known and unknown cyber threats.

As soon as the words "AI" and "music" are used in the same sentence, one comes across skepticism. If robots are making call centre jobs useless, one is scared to think of what would happen to all the musicans who are anyway underpaid. "In the world of personalisation and on-demand services, music is one of the very few remaining static artefacts," says Ken Lythgoe, head of business development at creative AI technology company MXX, based in London, England. The company has created the world's first AI tech that allows individual users to instantly edit music to fit their own video footage, complete with rises and fades. According to Lythgoe, AI doesn't need to be the enemy of music, and instead of replacing us, AI can empower us. MXX's AI tech listens to music and creates a metadata based on its understanding of it.

Computer pioneer Margaret Hamilton was critical to landing astronauts on the moon for the first time on 20 July 1969 and returning them safely a few days later. The young Massachusetts Institute of Technology (MIT) computer programmer and working mother led the team that created the onboard flight software for the Apollo missions, including Apollo 11. The computer system was the most sophisticated of its day. Her rigorous approach was so successful that no software bugs were ever known to have occurred during any crewed Apollo missions. "She symbolises that generation of unsung women who helped send humankind into space," said President Barack Obama in 2016 when he awarded Hamilton the Presidential Medal of Freedom, the United States' highest civilian award.