Smart grids enhance power grids by integrating electronic equipment, communication systems and computational tools. In a smart grid, consumers can insert energy into the power grid. We propose a new energy management system (called RLbEMS) that autonomously defines a policy for selling or storing energy surplus in smart homes. This policy is achieved through Batch Reinforcement Learning with historical data about energy prices, energy generation, consumer demand and characteristics of storage systems. In practical problems, RLbEMS has learned good energy selling policies quickly and effectively. We obtained maximum gains of 20.78% and 10.64%, when compared to a Naive-greedy policy, for smart homes located in Brazil and in the USA, respectively. Another important result achieved by RLbEMS was the reduction of about 30% of peak demand, a central desideratum for smart grids.

The Scaffolded Sound Beehive is an immersive multi-media installation which provides viewers an artistic visual and audio experience of activities in a beehive. Data were recorded in urban beehives and processed using sophisticated pattern recognition, AI technologies, and sonification and computer graphics software. The installation includes an experiment in using Deep Learning to interpret the activities in the hive based on sound and microclimate recording.

Nowadays, robots are programmed with a mix of discrete and continuous low level behaviors by experts in a very time consuming and expensive process. Existing automated planning approaches are either based on hybrid model predictive control techniques, which do not scale well due to time discretization, or temporal planners, which sacrifice plan expressivity by only supporting discretized fixed rates of change in continuous effects. We introduce Scotty, a mixed discrete-continuous generative planner that finds the middle ground between these two. Scotty can reason with linear time evolving effects whose behaviors can be modified by bounded control variables, with no discretization involved. Our planner exploits the expressivity of flow tubes, which compactly encapsulate continuous effects, and the performance of heuristic forward search. The generated solution plans are better suited for robust execution, as executives can use the flexibility in both time and continuous control variables to react to disturbances.

We consider discrete protocols for the classical Steinhaus cake cutting problem. Under mild technical conditions, we show that any deterministic strategy-proof protocol for two agents in the standard Robertson-Webb query model is dictatorial, that is, there is a fixed agent to which the protocol allocates the entire cake. For n > 2 agents, a similar impossibility holds, namely there always exists an agent that gets the empty piece (i.e. no cake). In contrast, we exhibit randomized protocols that are truthful in expectation and compute approximately fair allocations.

We study the adjustment and use of the Max-sumalgorithm for solving Asymmetric Distributed ConstraintOptimization Problems (ADCOPs). First, we formalize asymmetric factor-graphs and apply the different versions of Max-sum to them. Apparently, in contrast to local search algorithms, most Max-sum versions perform similarly when solving symmetric and asymmetric problems and some even perform better on asymmetric problems. Second, we prove that the convergence properties of Max-sum ADVP (an algorithm that was previously found to outperform other Max-sum versions) and the quality of the solutions it produces are dependent on the order between nodes involved in each constraint, i.e., the inner constraint order (ICO). A standard ICO allows to reproduce the properties achieved for symmetric problems, and outperform previously proposed local search ADCOP algorithms. Third, we demonstrate that a non-standard ICO can be used to balance exploration and exploitation, resulting in the best performing Max-sum version on both symmetric and asymmetric standard benchmarks.

Standard results on and algorithms for repeated games assume that defections are instantly observable. In reality, it may take some time for the knowledge that a defection has occurred to propagate through the social network. How does this affect the structure of equilibria and algorithms for computing them? In this paper, we consider games with cooperation and defection. We prove that there exists a unique maximal set of forever-cooperating agents in equilibrium and give an efficient algorithm for computing it. We then evaluate this algorithm on random graphs and find experimentally that there appears to be a phase transition between cooperation everywhere and defection everywhere, based on the value of cooperation and the discount factor. Finally, we provide a condition for when the equilibrium found is credible, in the sense that agents are in fact motivated to punish deviating agents. We find that this condition always holds in our experiments, provided the graphs are sufficiently large.

Exploiting parallelism is a key strategy for speeding up computation. However, on hard combinatorial problems, such a strategy has been surprisingly challenging due to the intricate variable interactions.We introduce a novel way in which parallelism can be used to exploit hidden structure of hard combinatorial problems. Our approach complements divide-and-conquer and portfolio approaches. We evaluate our approach on the minimum set basis problem: a core combinatorial problem with a range of applications in optimization, machine learning, and system security. We also highlight a novel sustainability related application, concerning the discovery of new materials for renewable energy sources such as improved fuel cell catalysts. In our approach, a large number of smaller sub-problems are identified and solved concurrently. We then aggregate the information from those solutions, and use this information to initialize the search of a global, complete solver. We show that this strategy leads to a substantial speed-up over a sequential approach, since the aggregated sub-problem solution information often provides key structural insights to the complete solver. Our approach also greatly outperforms state-of-the-art incomplete solvers in terms of solution quality. Our work opens up a novel angle for using parallelism to solve hard combinatorial problems.

Bayesian optimization is an effective methodology for the global optimization of functions with expensive evaluations. It relies on querying a distribution over functions defined by a relatively cheap surrogate model. An accurate model for this distribution over functions is critical to the effectiveness of the approach, and is typically fit using Gaussian processes (GPs). However, since GPs scale cubically with the number of observations, it has been challenging to handle objectives whose optimization requires many evaluations, and as such, massively parallelizing the optimization. In this work, we explore the use of neural networks as an alternative to GPs to model distributions over functions. We show that performing adaptive basis function regression with a neural network as the parametric form performs competitively with state-of-the-art GP-based approaches, but scales linearly with the number of data rather than cubically. This allows us to achieve a previously intractable degree of parallelism, which we apply to large scale hyperparameter optimization, rapidly finding competitive models on benchmark object recognition tasks using convolutional networks, and image caption generation using neural language models.

Given an overcomplete dictionary $A$ and a signal $b$ that is a linear combination of a few linearly independent columns of $A$, classical sparse recovery theory deals with the problem of recovering the unique sparse representation $x$ such that $b = A x$. It is known that under certain conditions on $A$, $x$ can be recovered by the Basis Pursuit (BP) and the Orthogonal Matching Pursuit (OMP) algorithms. In this work, we consider the more general case where $b$ lies in a low-dimensional subspace spanned by some columns of $A$, which are possibly linearly dependent. In this case, the sparsest solution $x$ is generally not unique, and we study the problem that the representation $x$ identifies the subspace, i.e. the nonzero entries of $x$ correspond to dictionary atoms that are in the subspace. Such a representation $x$ is called subspace-sparse. We present sufficient conditions for guaranteeing subspace-sparse recovery, which have clear geometric interpretations and explain properties of subspace-sparse recovery. We also show that the sufficient conditions can be satisfied under a randomized model. Our results are applicable to the traditional sparse recovery problem and we get conditions for sparse recovery that are less restrictive than the canonical mutual coherent condition. We also use the results to analyze the sparse representation based classification (SRC) method, for which we get conditions to show its correctness.

Machine learning methods for solving the equations of dynamical mean-field theory are developed. The method is demonstrated on the three dimensional Hubbard model. The key technical issues are defining a mapping of an input function to an output function, and distinguishing metallic from insulating solutions. Both metallic and Mott insulator solutions can be predicted. The validity of the machine learning scheme is assessed by comparing predictions of full correlation functions, of quasi-particle weight and particle density to values directly computed. The results indicate that with modest further development, machine learning approach may be an attractive computational efficient option for real materials predictions for strongly correlated systems.