Rich Rogers, a senior vice president of product and engineering at Hitachi Vantara, envisions a data center in which AI-driven management software (some or all of it cloud-based) will monitor and control IT and facilities infrastructure, as well as applications, seamlessly and completely across single or multiple sites. Compute, power, storage, networking and cooling operations will flex dynamically to achieve maximum efficiency, productivity and availability. Human operators, meanwhile, will be free to do what they do best: plan new capabilities and innovate improvements. "IoT and AI will enable data center issues to be root-caused and resolved automatically by software," Rogers said. Data center administrators will no longer be woken-up at night to troubleshoot outages.

Zoning codes are a century old, and the lifeblood of all major U.S. cities (except arguably Houston), determining what can be built where and what activities can take place in a neighborhood. Yet as their complexity has risen, academics are increasingly exploring whether their rule-based systems for rationalizing urban space could be replaced with dynamic systems based on blockchains, machine learning algorithms, and spatial data, potentially revolutionizing urban planning and development for the next one hundred years. These visions of the future were inspired by my recent chats with Kent Larson and John Clippinger, a dynamic urban thinking duo who have made improving cities and urban governance their current career focus. Larson is a principal research scientist at the MIT Media Lab, where he directs the City Science Group, and Clippinger is a visiting researcher at the Human Dynamics Lab (also part of the Media Lab), as well as the founder of non-profit ID3. One of the toughest challenges facing major U.S. cities is the price of housing, which has skyrocketed over the past few decades, placing incredible strain on the budget of young and old, singles and families alike.

Just as the Industrial Revolution transformed the world during the 18th and 19th centuries, we are facing the dawn of an equally far-reaching artificial intelligence or AI revolution that will be measured in years rather than decades. AI has reached the point where it is capable of surpassing the decision-making of humans in many situations; consistently, accurately, 24/7 and based on more facts. But why now and how do businesses harness this capability? AI has been around since the 1960s; only now is the confluence of three key factors coming together. Firstly, AI has languished as a "disembodied brain in a jar", isolated from the real world.

In December, Oakland-based U.S. District Judge Haywood Gilliam blocked the changes to President Barack Obama's health care law. The judge said the Trump administration failed to give the notice and comment period required before implementing the changes.

We've been promised so much over the years, but now AI is ready to deliver. From connected cars and smart homes, through to improved ambient assisted living (AAL), AI is the common thread that unites them all. Analysts predicted that worldwide revenues for cognitive and artificial intelligence (AI) systems will reach $12.5 billion in 2017, an increase of 59.3% over 2016, and that investment will continue that trajectory, achieving a compound annual growth rate (CAGR) of 54.4% through 2020 when revenues will be more than $46 billion. The analysts might just be right this time too, as the supporting base of technology has caught up with the demands and opportunities offered by improved AI. A clear example was at CES 2018, one of the biggest annual showcases of technology, which lit up Las Vegas in early January, and much of the major noise this year was around AI and associated devices.

Interpretable machine learning models have received increasing interest in recent years, especially in domains where humans are involved in the decision-making process. However, the possible loss of the task performance for gaining interpretability is often inevitable. This performance downgrade puts practitioners in a dilemma of choosing between a top-performing black-box model with no explanations and an interpretable model with unsatisfying task performance. In this work, we propose a novel framework for building a Hybrid Decision Model that integrates an interpretable model with any black-box model to introduce explanations in the decision making process while preserving or possibly improving the predictive accuracy. We propose a novel metric, explainability, to measure the percentage of data that are sent to the interpretable model for decision. We also design a principled objective function that considers predictive accuracy, model interpretability, and data explainability. Under this framework, we develop Collaborative Black-box and RUle Set Hybrid (CoBRUSH) model that combines logic rules and any black-box model into a joint decision model. An input instance is first sent to the rules for decision. If a rule is satisfied, a decision will be directly generated. Otherwise, the black-box model is activated to decide on the instance. To train a hybrid model, we design an efficient search algorithm that exploits theoretically grounded strategies to reduce computation. Experiments show that CoBRUSH models are able to achieve same or better accuracy than their black-box collaborator working alone while gaining explainability. They also have smaller model complexity than interpretable baselines.

Consider a number of wind turbines scattered throughout the North Sea. Each turbine is equipped with a Query processing over data streams is a key aspect of Big sensor, which continuously records temperature levels of key Data applications. For instance, algorithmic trading relies on devices within the turbine and sends those readings to a data real-time analysis of stock tickers and financial news items centre monitoring the functioning of the turbines. Temperature (Nuti et al. 2011); oil and gas companies continuously monitor levels are streamed by sensors using a ternary predicate and analyse data coming from their wellsites in order Temp, whose arguments identify the device, the temperature to detect equipment malfunction and predict maintenance level, and the time of the reading. A monitoring task in the needs (Cosad et al. 2009); network providers perform realtime data centre is to track the activation of cooling measures in analysis of network flow data to identify traffic anomalies each turbine, record temperature-induced malfunctions and and DoS attacks (Münz and Carle 2007).

Existential rules, a family of expressive ontology languages, inherit desired expressive and reasoning properties from both description logics and logic programming. On the other hand, forgetting is a well studied operation for ontology reuse, obfuscation and analysis. Yet it is challenging to establish a theory of forgetting for existential rules. In this paper, we lay the foundation for a theory of forgetting for existential rules by developing a novel notion of unfolding. In particular, we introduce a definition of forgetting for existential rules in terms of query answering and provide a characterisation of forgetting by the unfolding. A result of forgetting may not be expressible in existential rules, and we then capture the expressibility of forgetting by a variant of boundedness. While the expressibility is undecidable in general, we identify a decidable fragment. Finally, we provide an algorithm for forgetting in this fragment.

We introduce a novel model-agnostic system that explains the behavior of complex models with high-precision rules called anchors, representing local, "sufficient" conditions for predictions. We propose an algorithm to efficiently compute these explanations for any black-box model with high-probability guarantees. We demonstrate the flexibility of anchors by explaining a myriad of different models for different domains and tasks. In a user study, we show that anchors enable users to predict how a model would behave on unseen instances with less effort and higher precision, as compared to existing linear explanations or no explanations.

In machine learning, rule models are one of the most popular choices when model interpretability is the primary concern. Ordinary, a single model is obtained by solving an optimization problem, and the resulting model is interpreted as the one that best explains the data. In this study, instead of finding a single rule model, we propose algorithms for enumerating multiple rule models. Model enumeration is useful in practice when (i) users want to choose a model that is particularly suited to their task knowledge, or (ii) users want to obtain several possible mechanisms that could be underlying the data to use as hypotheses for further scientific studies. To this end, we propose two enumeration algorithms: an approximate algorithm and an exact algorithm. We prove that these algorithms can enumerate models in a descending order of their objective function values approximately and exactly. We then confirm our theoretical results through experiments on real-world data. We also show that, by using the proposed enumeration algorithms, we can find several different models of almost equal quality.