The problem of keyword spotting i.e. identifying keywords in a real-time audio stream is mainly solved by applying a neural network over successive sliding windows. Due to the difficulty of the task, baseline models are usually large, resulting in a high computational cost and energy consumption level. We propose a new method called SANAS (Stochastic Adaptive Neural Architecture Search) which is able to adapt the architecture of the neural network on-the-fly at inference time such that small architectures will be used when the stream is easy to process (silence, low noise, ...) and bigger networks will be used when the task becomes more difficult. We show that this adaptive model can be learned end-to-end by optimizing a trade-off between the prediction performance and the average computational cost per unit of time. Experiments on the Speech Commands dataset show that this approach leads to a high recognition level while being much faster (and/or energy saving) than classical approaches where the network architecture is static.

Non-intrusive load monitoring (NILM), also known as energy disaggregation, is a blind source separation problem where a household's aggregate electricity consumption is broken down into electricity usages of individual appliances. In this way, the cost and trouble of installing many measurement devices over numerous household appliances can be avoided, and only one device needs to be installed. The problem has been well-known since Hart's seminal paper in 1992, and recently significant performance improvements have been achieved by adopting deep networks. In this work, we focus on the idea that appliances have on/off states, and develop a deep network for further performance improvements. Specifically, we propose a subtask gated network that combines the main regression network with an on/off classification subtask network. Unlike typical multitask learning algorithms where multiple tasks simply share the network parameters to take advantage of the relevance among tasks, the subtask gated network multiply the main network's regression output with the subtask's classification probability. When standby-power is additionally learned, the proposed solution surpasses the state-of-the-art performance for most of the benchmark cases. The subtask gated network can be very effective for any problem that inherently has on/off states.

Despite offering vast potential for energy efficiency, artificial intelligence and the Internet of Things generate widely divergent views among commercial real estate professionals. In the first part of a special two-part interview, Yardi Energy's Christy Cannon takes stock of how the IoT and AI are shaping property operations. By 2008, there were already more IoT devices than people on the planet, and that growth is continuing exponentially. It fits nicely with cloud computing and artificial intelligence like a fine wine, cheese, and fruit plate. It doesn't hurt that the cost of these little devices has decreased substantially over the years, and that the evolution of artificial intelligence utilizing IoT has enabled a building operator to efficiently balance comfort with energy and utilize space like never before.

Directed exploration strategies for reinforcement learning are critical for learning an optimal policy in a minimal number of interactions with the environment. Many algorithms use optimism to direct exploration, either through visitation estimates or upper confidence bounds, as opposed to data-inefficient strategies like \epsilon-greedy that use random, undirected exploration. Most data-efficient exploration methods require significant computation, typically relying on a learned model to guide exploration. Least-squares methods have the potential to provide some of the data-efficiency benefits of model-based approaches -- because they summarize past interactions -- with the computation closer to that of model-free approaches. In this work, we provide a novel, computationally efficient, incremental exploration strategy, leveraging this property of least-squares temporal difference learning (LSTD). We derive upper confidence bounds on the action-values learned by LSTD, with context-dependent (or state-dependent) noise variance. Such context-dependent noise focuses exploration on a subset of variable states, and allows for reduced exploration in other states. We empirically demonstrate that our algorithm can converge more quickly than other incremental exploration strategies using confidence estimates on action-values.

Material irradiation experiment is dangerous and complex, thus it requires those with a vast advanced expertise to process the images and data manually. In this paper, we propose a generative adversarial model based on prior knowledge and attention mechanism to achieve the generation of irradiated material images (data-to-image model), and a prediction model for corresponding industrial performance (image-to-data model). With the proposed models, researchers can skip the dangerous and complex irradiation experiments and obtain the irradiation images and industrial performance parameters directly by inputing some experimental parameters only. We also introduce a new dataset ISMD which contains 22000 irradiated images with 22,143 sets of corresponding parameters. Our model achieved high quality results by compared with several baseline models. The evaluation and detailed analysis are also performed.

We suggest a new methodology for analysis and computations that combines the gauge transformation (GT) technique from (Chertkov, Chernyak 2006) with the technique developed in (Gurvits 2011, Anari, Gharan 2017, Straszak, Vishnoi 2017) based on the recent progress in the field of real stable polynomials. We show that GTs (while keeping PF invariant) allow representation of PF as a sum of polynomials of variables associated with edges of the graph. A special belief propagation (BP) gauge makes a single out term of the series least sensitive to variations then resulting in the loop series for PF introduced in (Chertkov, Chernyak 2006). In addition to restating the known results in the polynomial form, we also discover a new relation between the computationally tractable BP term (single out term of the loop series evaluated at the BP gauge) and the PF: sequential application of differential operators, each associated with an edge of the graph, to the BP polynomial results in the PF. Each term in the sequence corresponds to a BP polynomial of a modified GM derived by contraction of an edge. Even though complexity of computing factors in the derived GMs grow exponentially with the number of eliminated edges, polynomials associated with the new factors remain bi-stable if the original factors have this property. Moreover, we show that BP estimations for the PF do not decrease with eliminations, thus resulting overall in a new proof of the result following from a combination of (Anari, Gharan 2017) and (Straszak, Vishnoi 2017) that the BP solution of the original GM with factors correspondent to bi-stable polynomials gives a lower bound for PF.

In this paper, we introduce proximal gradient temporal difference learning, which provides a principled way of designing and analyzing true stochastic gradient temporal difference learning algorithms. We show how gradient TD (GTD) reinforcement learning methods can be formally derived, not by starting from their original objective functions, as previously attempted, but rather from a primal-dual saddle-point objective function. We also conduct a saddle-point error analysis to obtain finite-sample bounds on their performance. Previous analyses of this class of algorithms use stochastic approximation techniques to prove asymptotic convergence, and do not provide any finite-sample analysis. We also propose an accelerated algorithm, called GTD2-MP, that uses proximal "mirror maps" to yield an improved convergence rate. The results of our theoretical analysis imply that the GTD family of algorithms are comparable and may indeed be preferred over existing least squares TD methods for off-policy learning, due to their linear complexity. We provide experimental results showing the improved performance of our accelerated gradient TD methods.

We apply numerical methods in combination with finite-difference-time-domain (FDTD) simulations to optimize transmission properties of plasmonic mirror color filters using a multi-objective figure of merit over a five-dimensional parameter space by utilizing novel multi-fidelity Gaussian processes approach. We compare these results with conventional derivative-free global search algorithms, such as (single-fidelity) Gaussian Processes optimization scheme, and Particle Swarm Optimization---a commonly used method in nanophotonics community, which is implemented in Lumerical commercial photonics software. We demonstrate the performance of various numerical optimization approaches on several pre-collected real-world datasets and show that by properly trading off expensive information sources with cheap simulations, one can more effectively optimize the transmission properties with a fixed budget.

In machine learning, a nonparametric forecasting algorithm for time series data has been proposed, called the kernel spectral hidden Markov model (KSHMM). In this paper, we propose a technique for short-term wind-speed prediction based on KSHMM. We numerically compared the performance of our KSHMMbased forecasting technique to other techniques with machine learning, using wind-speed data offered by the National Renewable Energy Laboratory. Our results demonstrate that, compared to these methods, the proposed technique offers comparable or better performance. Keywords: Wind-Speed Prediction, Kernel Methods, Kernel Mean Embedding, Spectral Learning, Hidden Markov Models. 1. Introduction Wind energy is one of the most attractive renewable energy sources.

Matter evolved under influence of gravity from minuscule density fluctuations. Non-perturbative structure formed hierarchically over all scales, and developed non-Gaussian features in the Universe, known as the Cosmic Web. To fully understand the structure formation of the Universe is one of the holy grails of modern astrophysics. Astrophysicists survey large volumes of the Universe and employ a large ensemble of computer simulations to compare with the observed data in order to extract the full information of our own Universe. However, to evolve trillions of galaxies over billions of years even with the simplest physics is a daunting task. We build a deep neural network, the Deep Density Displacement Model (hereafter D$^3$M), to predict the non-linear structure formation of the Universe from simple linear perturbation theory. Our extensive analysis, demonstrates that D$^3$M outperforms the second order perturbation theory (hereafter 2LPT), the commonly used fast approximate simulation method, in point-wise comparison, 2-point correlation, and 3-point correlation. We also show that D$^3$M is able to accurately extrapolate far beyond its training data, and predict structure formation for significantly different cosmological parameters. Our study proves, for the first time, that deep learning is a practical and accurate alternative to approximate simulations of the gravitational structure formation of the Universe.