Phase segregation, the process by which the components of a binary mixture spontaneously separate, is a key process in the evolution and design of many chemical, mechanical, and biological systems. In this work, we present a data-driven approach for the learning, modeling, and prediction of phase segregation. A direct mapping between an initially dispersed, immiscible binary fluid and the equilibrium concentration field is learned by conditional generative convolutional neural networks. Concentration field predictions by the deep learning model conserve phase fraction, correctly predict phase transition, and reproduce area, perimeter, and total free energy distributions up to 98% accuracy.

The 4th Industrial Revolution continues to demonstrate what I call exponential "A Triple C": Perhaps nowhere is this "A Triple C" dynamic more on display than in the realm of artificial intelligence (AI), which is expected to underpin many of the key emerging technologies and power business growth across industries. It's clear that AI is becoming the new electricity, and its rapid proliferation has happened in a very short span of three years – with a large leap in the past 12 months. But what's really exciting is AI's potential to improve lives at a pace and scale not seen before. Aiding this potential is a significant business and investment shift toward a greater focus on social good. Taken together, these dynamics are now resulting in a rising number of use cases for the application of AI to accelerate progress on the United Nations' Sustainable Development Goals (SDGs).

When I began writing my book "The Deep Learning A.I. Playbook", I had given very little thought about the dangers of Artificial Intelligence (AI). I was fortunate however to be able to form a bit of an understanding to write a chapter about Human Compatible AI (A term Stuart Russell uses to frame the problem) as a bookend for my book. I have however, come to the realization that the A.I. problem is a problem that is inextricably intertwined with human civilization. It cannot be solved because present human civilization isn't structured in a manner that is aligned with the needs of humanity. You cannot achieve human beneficial A.I. without drastically remaking human civilization.

NEW YORK – For a dozen students from Futaba Future High School in Fukushima Prefecture, a recent visit to the United Nations was a chance to share their plans to improve the lives of others by drawing from their catastrophic earthquake and tsunami experiences as a source of strength. Despite overcoming enormous hurdles in the aftermath of the March 11, 2011, disaster that took more than 19,000 lives, the surviving students have moved forward with aspirations of choosing future paths to benefit the global community. "Thanks to all my experiences like getting bullied, joining the drama club and studying at my high school, I think I could grow well," Satsuki Sekine told U.N. diplomats, staff and youth representatives who gathered to hear their presentation on the current situation in Fukushima early this month as part of a scheduled visit while in New York. The 17-year-old explained how drama can be used to portray the challenges of discrimination and conflict "not as an abstract concept but with specific and visual examples." Recounting how the tsunami rendered her home unlivable, she explained how her life in Tomioka as a normal 9-year-old was turned upside down.

The commoditization of trust has been the topic of science fiction, futuristic novels and theoretical study for the last century. Advances in blockchain and artificial intelligence technology continue to make science-fiction a reality, automating and replacing the need for 3rd party intermediaries and other trust mechanisms, potentially disrupting many critical industries. Blockchain enabled smart contracts show potential to exchange value without third party trust mechanisms. The combination of artificial intelligence, cryptography, distributed trust algorithms or smart contracts have paved the way to a more efficient and secure way to exchange value, goods and services. This paper explores how blockchain technology could potentiality automate and modernize energy and the internet of things to help evolve energy infrastructure to an increasingly automated, distributed, clean and resilient system. This is timely as the U.S. power grid and the array of things that it connects to is a complex system of systems in which the nation’s economy, national security and livelihood depends on.

In a recent paper, the authors proposed a general methodology for probabilistic learning on manifolds. The method was used to generate numerical samples that are statistically consistent with an existing dataset construed as a realization from a non-Gaussian random vector. The manifold structure is learned using diffusion manifolds and the statistical sample generation is accomplished using a projected Ito stochastic differential equation. This probabilistic learning approach has been extended to polynomial chaos representation of databases on manifolds and to probabilistic nonconvex constrained optimization with a fixed budget of function evaluations. The methodology introduces an isotropic-diffusion kernel with hyperparameter {\epsilon}. Currently, {\epsilon} is more or less arbitrarily chosen. In this paper, we propose a selection criterion for identifying an optimal value of {\epsilon}, based on a maximum entropy argument. The result is a comprehensive, closed, probabilistic model for characterizing data sets with hidden constraints. This entropy argument ensures that out of all possible models, this is the one that is the most uncertain beyond any specified constraints, which is selected. Applications are presented for several databases.

In this paper, we consider a realistic and meaningful scenario in the context of smart grids where an electricity retailer serves three different types of customers, i.e., customers with an optimal home energy management system embedded in their smart meters (C-HEMS), customers with only smart meters (C-SM), and customers without smart meters (C-NONE). The main objective of this paper is to support the retailer to make optimal day-ahead dynamic pricing decisions in such a mixed customer pool. To this end, we propose a two-level decision-making framework where the retailer acting as upper-level agent firstly announces its electricity prices of next 24 hours and customers acting as lower-level agents subsequently schedule their energy usages accordingly. For the upper level problem, we optimize the dynamic prices for the retailer to maximize its profit subject to realistic market constraints. The above two-level model is tackled by genetic algorithms (GA) based distributed optimization methods while its feasibility and effectiveness are con-2018. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/. Please cite this accepted article as: Fanlin Meng, Xiao-Jun Zeng, Yan Zhang, Chris J. Dent, Dunwei Gong, An Integrated Optimization Learning Approach to Optimal Dynamic Pricing for the Retailer with Multi-type Customers in Smart Grids, Information Sciences (2018), doi: 10.1016/j.ins.2018.03.039 Preprint submitted to Information Sciences March 22, 2018 firmed via simulation results. Keywords: Bilevel Modelling, Genetic Algorithms, Machine Learning, Dynamic Pricing, Demand-side Management, Demand Response, Smart Grids 1. Introduction With the large-scale deployment of smart meters and two-way communication infrastructures, dynamic pricing based demand response and demand-side management programs [37] [12] have attracted enormous attentions from both academia and industry and are expected to bring great benefits to the whole power system. Real-time pricing (RTP), timeof-use pricing (ToU) and critical-peak pricing (CPP) are commonly used dynamic pricing strategies [20].

You return home to your penthouse apartment after a long day at work auctioning Cryptokitties and other cryptogoods on a peer-to-peer marketplace. You grab a bottle of tangerine-flavored weed soda from the fridge and sink into your couch. A wooden side table, custom-built by a robot in India, holds a box containing your antidepressant patches. You peel off the back and slap one on your arm. The television replies in the warm, fully human-sounding female whisper you chose for it: "We've got five new videos from your favorite ASMR channels.

A useful computation when acting in a complex environment is to infer the marginal probabilities or most probable states of task-relevant variables. Probabilistic graphical models can efficiently represent the structure of such complex data, but performing these inferences is generally difficult. Message-passing algorithms, such as belief propagation, are a natural way to disseminate evidence amongst correlated variables while exploiting the graph structure, but these algorithms can struggle when the conditional dependency graphs contain loops. Here we use Graph Neural Networks (GNNs) to learn a message-passing algorithm that solves these inference tasks. We first show that the architecture of GNNs is well-matched to inference tasks. We then demonstrate the efficacy of this inference approach by training GNNs on an ensemble of graphical models and showing that they substantially outperform belief propagation on loopy graphs. Our message-passing algorithms generalize out of the training set to larger graphs and graphs with different structure.

Having once been an early adopter of tech, I was an unlikely early rejector. But it has now been over a year since I have phoned my family or friends, logged on to antisocial media, sent a text message, checked email, browsed online, took a photograph or listened to electronic music. Living and working on a smallholding without electricity, fossil fuels or running water, the last year has taught me much about the natural world, society, the state of our shared culture, and what it means to be human in a time when the boundaries between man and machine are blurring. My reasons for unplugging, during that time, haven't so much changed as shifted in importance. My primary motives were – and still are – ecological.