The conjunction of psyche and matter, of digital and physical advances, lies at the core of the fourth industrial revolution. The marriage of Artificial Intelligence (AI) and materials science speaks as one of the clearest models. Unadulterated digital development has pulled in the best consideration and a large offer of financial investment in the course of the most recent years. Be that as it may, we live in a material world, where the nature of our lives relies upon enhancements in physical products and services: nourishment and asylum, social insurance, transportation, energy etc. It is quite true that we invest much more energy in our online virtual universes, yet this is reflected by a developing number of Amazon bundles at our doorsteps.

Hierarchical text classification, which aims to classify text documents into a given hierarchy, is an important task in many real-world applications. Recently, deep neural models are gaining increasing popularity for text classification due to their expressive power and minimum requirement for feature engineering. However, applying deep neural networks for hierarchical text classification remains challenging, because they heavily rely on a large amount of training data and meanwhile cannot easily determine appropriate levels of documents in the hierarchical setting. In this paper, we propose a weakly-supervised neural method for hierarchical text classification. Our method does not require a large amount of training data but requires only easy-to-provide weak supervision signals such as a few class-related documents or keywords. Our method effectively leverages such weak supervision signals to generate pseudo documents for model pre-training, and then performs self-training on real unlabeled data to iteratively refine the model. During the training process, our model features a hierarchical neural structure, which mimics the given hierarchy and is capable of determining the proper levels for documents with a blocking mechanism. Experiments on three datasets from different domains demonstrate the efficacy of our method compared with a comprehensive set of baselines.

Machine learning systems increasingly depend on pipelines of multiple algorithms to provide high quality and well structured predictions. This paper argues interaction effects between clustering and prediction (e.g. classification, regression) algorithms can cause subtle adverse behaviors during cross-validation that may not be initially apparent. In particular, we focus on the problem of estimating the out-of-cluster (OOC) prediction loss given an approximate clustering with probabilistic error rate $p_0$. Traditional cross-validation techniques exhibit significant empirical bias in this setting, and the few attempts to estimate and correct for these effects are intractable on larger datasets. Further, no previous work has been able to characterize the conditions under which these empirical effects occur, and if they do, what properties they have. We precisely answer these questions by providing theoretical properties which hold in various settings, and prove that expected out-of-cluster loss behavior rapidly decays with even minor clustering errors. Fortunately, we are able to leverage these same properties to construct hypothesis tests and scalable estimators necessary for correcting the problem. Empirical results on benchmark datasets validate our theoretical results and demonstrate how scaling techniques provide solutions to new classes of problems.

Objective: A patient medical insurance coverage plays an essential role in determining the post-acute care (PAC) discharge disposition. The prior health insurance authorization process postpones the PAC discharge disposition, increases the inpatient length of stay, and effects patient health. Our study implements predictive analytics for the early prediction of the PAC discharge disposition to reduce the deferments caused by prior health insurance authorization, the inpatient length of stay and inpatient stay expenses. Methodology: We conducted a group discussion involving 25 patient care facilitators (PCFs) and two registered nurses (RNs) and retrieved 1600 patient data records from the initial nursing assessment and discharge notes to conduct a retrospective analysis of PAC discharge dispositions using predictive analytics. Results: The chi-squared automatic interaction detector (CHAID) algorithm enabled the early prediction of the PAC discharge disposition, accelerated the prior health insurance process, decreased the inpatient length of stay by an average of 22.22%, and reduced inpatient stay expenses by \$1,974 for state government hospitals, \$2,346 for non-profit hospitals and \$1,798 for for-profit hospitals per day. The CHAID algorithm produced an overall accuracy of 84.16% and an area under the receiver operating characteristic (ROC) curve value of 0.81. Conclusion: The early prediction of PAC discharge dispositions can condense the PAC deferment caused by the prior health insurance authorization process and simultaneously minimize the inpatient length of stay and related expenses incurred by the hospital.

Understanding the phenotypic drug response on cancer cell lines plays a vital rule in anti-cancer drug discovery and re-purposing. The Genomics of Drug Sensitivity in Cancer (GDSC) database provides open data for researchers in phenotypic screening to test their models and methods. Previously, most research in these areas starts from the fingerprints or features of drugs, instead of their structures. In this paper, we introduce a model for phenotypic screening, which is called twin Convolutional Neural Network for drugs in SMILES format (tCNNS). tCNNS is comprised of CNN input channels for drugs in SMILES format and cancer cell lines respectively. Our model achieves $0.84$ for the coefficient of determinant($R^2$) and $0.92$ for Pearson correlation($R_p$), which are significantly better than previous works\cite{ammad2014integrative,haider2015copula,menden2013machine}. Besides these statistical metrics, tCNNS also provides some insights into phenotypic screening.

Honey has been collected and used by humankind as both a food and medicine for thousands of years. However, in the modern economy, honey has become subject to mislabelling and adulteration making it the third most faked food product in the world. The international scale of fraudulent honey has had both economic and environmental ramifications. In this paper, we propose a novel method of identifying fraudulent honey using machine learning augmented microscopy.

Much of what we hear about artificial intelligence and machine learning in security products is steeped in marketing, making it hard to know what these tools actually do.

Microsoft Corp. called for new legislation to govern artificial intelligence software for recognizing faces, advocating for human review and oversight of the technology in critical cases. "This includes where decisions may create a risk of bodily or emotional harm to a consumer, where there may be implications on human or fundamental rights, or where a consumer's personal freedom or privacy may be impinged," Microsoft President and Chief Legal Officer Brad Smith wrote in a blog published in conjunction with a speech on the topic at the Brookings Institution think tank. Sellers of the technology must "recognize that they are not absolved of their obligation to comply with laws prohibiting discrimination against individual consumers or groups of consumers," he added. Smith also wants laws to require sellers of the products to explain what they do clearly and open up their services to testing by outside parties for accuracy and bias. Earlier Thursday, advocacy group AI Now called for greater regulation and regular audits of AI tools used by governments.

Google's first foray into Chinese markets was a short-lived experiment. Google China's search engine was launched in 2006 and abruptly pulled from mainland China in 2010 amid a major hack of the company and disputes over censorship of search results. But in August 2018, the investigative journalism website The Intercept reported that the company was working on a secret prototype of a new, censored Chinese search engine, called Project Dragonfly. Amid a furor from human rights activists and some Google employees, US Vice President Mike Pence called on the company to kill Dragonfly, saying it would "strengthen Communist Party censorship and compromise the privacy of Chinese customers." In mid-December, The Intercept reported that Google had suspended its development efforts in response to complaints from the company's own privacy team, who learned about the project from the investigative website's reporting. Observers talk as if the decision about whether to reenter the world's largest market is up to Google: will it compromise its principles and censor search the way China wants?

Note: The below article includes excerpts from the September 2018 Health Affairs article "Implementing and Scaling Artificial Intelligence Solutions: Considerations for Policy Makers and Decision Makers." The full article can be accessed through the Health Affairs Journal, and was developed in partnership with Eric Just, MS, from Health Catalyst; and Bruce L. Gillingham, MD, CPE, FAOA RADM, MC, USN, Rear Admiral in the US Navy. The New England Journal of Medicine, Harvard Business Review, and other publications, noted that Artificial Intelligence (AI) and machine learning can achieve breakthroughs in improving patient safety, health, and reducing waste. AI infers patterns, relationships, and rules directly from large volumes of data in ways that can exceed human cognitive capabilities. Today, opportunities for greater deployment of AI in health are made possible by increased electronic data availability from mobile devices, sensors, cameras, and electronic health records; faster data processing capabilities; and newly developed computing techniques.