Many efficient algorithms with strong theoretical guarantees have been proposed for the contextual multi-armed bandit problem. However, applying these algorithms in practice can be difficult because they require domain expertise to build appropriate features and to tune their parameters. We propose a new method for the contextual bandit problem that is simple, practical, and can be applied with little or no domain expertise. Our algorithm relies on decision trees to model the context-reward relationship. Decision trees are non-parametric, interpretable, and work well without hand-crafted features. To guide the exploration-exploitation trade-off, we use a bootstrapping approach which abstracts Thompson sampling to non-Bayesian settings. We also discuss several computational heuristics and demonstrate the performance of our method on several datasets.

There has recently been an increased emphasis on reducing energy consumption in manufacturing, driven by the fluctuations in energy costs and the growing importance given to environmental impact of manufactured goods. Lots of attention has been given to the reduction of machine tools energy consumption, as they require large amounts of energy to perform manufacturing tasks. One area that has received relatively little interest, yet could harness great potential, is reducing energy consumption by planning machine activities between manufacturing operations, while the machine is not in use. The intuitive option --which is currently exploited in manufacturing-- is to leave the machine in a normal operating state in anticipation of the next manufacturing job. However, this is far from optimal due to the thermal deformation phenomenon, which usually require an energy-intensive warm-up cycle in order to bring all the components (e.g. spindle motor) into a suitable (stable) state for actual machining. Evidently, the use of this strategy comes with the associated commercial and environmental repercussions. In this paper, we investigate the exploitability of automated planning techniques for planning machine activities between manufacturing operations. We present a PDDL 2.2 formulation of the task that considers energy consumption, thermal deformation, and accuracy. We then demonstrate the effectiveness of the proposed approach using a case study which considers real-world data.

Energy costs are an increasingly important issue in real-world scheduling, for both economic and environmental reasons. This paper deals with a variant of the well-known job shop scheduling problem, where we consider a bi-objective optimization of both the weighted tardiness and the energy costs. To this end, we design a hybrid metaheuristic that combines a genetic algorithm with a novel local search method and a linear programming approach. We also propose an efficient procedure for improving the energy cost of a given schedule. In the experimental study we analyse our proposal and compare it with the state of the art and also with a constraint programming approach, obtaining competitive results.

Most applications of planning to real problems involve complex and often non-linear equations, including matrix operations. PDDL is ill-suited to express such calculations since it only allows basic operations between numeric fluents. To remedy this restriction, a generic PDDL planner can be connected to a specialised advisor, which equips the planner with the ability to carry out sophisticated mathematical operations. Unlike related techniques based on semantic attachment, our planner is able to exploit an approximation of the numeric information calculated by the advisor to compute informative heuristic estimators. Guided by both causal and numeric information, our planning framework outperforms traditional approaches, especially against problems with numeric goals. We provide evidence of the power of our solution by successfully solving four completely different problems.

Read through some examples of how the oil and gas industry will benefit from the combination of Artificial Intelligence and mobile development. To read the Bloomberg article about Apple's Artificial Intelligence chip, click here.

Franklin Robotics has launched a Kickstarter campaign for Tertill, their solar-powered, garden-weeding robot. Tertill lives in your garden, collecting sunlight to power its weed patrol, and cutting down short plants with a string trimmer/weed whacker with almost no intervention required. Available for about $300USD, the fully autonomous Tertill is the first weeding robot available to home gardeners. Tertill is round, short, has four wheel drive and extreme camber wheels. It uses proprietary algorithms to ensure that it finds as many weeds as it can, using its sensors to distinguish between weeds and crops based on height.

BERLIN--(BUSINESS WIRE)--Today at GE's Minds Machines Europe event, GE Ventures (NYSE:GE) announced the launch of Avitas Systems, a new company that will use predictive data analytics, robotics, and artificial intelligence to deliver advanced inspection services to the oil and gas, transportation, and energy industries. Routine inspections can be slow and costly, and often include humans performing high-risk tasks. Data is manually collected and processed, and can take weeks to analyze. By reducing high-risk tasks through robotics, Avitas Systems can make inspection processes safer and more efficient through data automation, decreasing costs by up to 25%. By performing inspections based on anticipated risk, instead of regular time intervals, Avitas Systems can also help to increase asset longevity.

We consider the problem of approximate Bayesian parameter inference in non-linear state-space models with intractable likelihoods. Sequential Monte Carlo with approximate Bayesian computations (SMC-ABC) is one approach to approximate the likelihood in this type of models. However, such approximations can be noisy and computationally costly which hinders efficient implementations using standard methods based on optimisation and Monte Carlo methods. We propose a computationally efficient novel method based on the combination of Gaussian process optimisation and SMC-ABC to create a Laplace approximation of the intractable posterior. We exemplify the proposed algorithm for inference in stochastic volatility models with both synthetic and real-world data as well as for estimating the Value-at-Risk for two portfolios using a copula model. We document speed-ups of between one and two orders of magnitude compared to state-of-the-art algorithms for posterior inference.

We provide a bridge between generative modeling in the Machine Learning community and simulated physical processes in High Energy Particle Physics by applying a novel Generative Adversarial Network (GAN) architecture to the production of jet images -- 2D representations of energy depositions from particles interacting with a calorimeter. We propose a simple architecture, the Location-Aware Generative Adversarial Network, that learns to produce realistic radiation patterns from simulated high energy particle collisions. The pixel intensities of GAN-generated images faithfully span over many orders of magnitude and exhibit the desired low-dimensional physical properties (i.e., jet mass, n-subjettiness, etc.). We shed light on limitations, and provide a novel empirical validation of image quality and validity of GAN-produced simulations of the natural world. This work provides a base for further explorations of GANs for use in faster simulation in High Energy Particle Physics.

Project Sunroof was launched in 2015 by Carl Elkin, an engineer at Google who had worked on local solar-installation campaigns in Massachusetts. It now provides data for 60 million homes across the United States that it has already assessed with its algorithms. For the past two years, Project Sunroof has walked people through all the information-gathering steps of installing solar panels: After you tell it where you live, its algorithms estimate how much solar energy falls on your roof, calculate how much solar panels would reduce your electricity bill, and deliver estimates from local installation firms like Solar City. It can also walk you through similar steps if you're interested in leasing or borrowing panels. "It highlights that, for many people, solar is often free. In many cases, including for my house, solar is better than free," Elkin told me last week.