14 January 2019 in Nature Genetics
29 March 2024 in Science Advances
Uncertainty quantification is crucial to decision-making. A prominent example is probabilistic forecasting in numerical weather prediction. The dominant approach to representing uncertainty in weather forecasting is to generate an ensemble of forecasts by running physics-based simulations under different conditions, which is a computationally costly process. We propose to amortize the computational cost by emulating these forecasts with deep generative diffusion models learned from historical data. The learned models are highly scalable with respect to high-performance computing accelerators and can sample thousands of realistic weather forecasts at low cost. When designed to emulate operational ensemble forecasts, the generated ones are similar to physics-based ensembles in statistical properties and predictive skill. When designed to correct biases present in the operational forecasting system, the generated ensembles show improved probabilistic forecast metrics. They are more reliable and forecast probabilities of extreme weather events more accurately. While we focus on weather forecasting, this methodology may enable creating large climate projection ensembles for climate risk assessment.
26 January 2010 in IEEE Transactions on Image Processing
In this correspondence, we propose a novel, efficient, and effective Refined Histogram (RH) for modeling the wavelet subband detail coefficients and present a new image signature based on the RH model for supervised texture classification. Our RH makes use of a step function with exponentially increasing intervals to model the histogram of detail coefficients, and the concatenation of the RH model parameters for all wavelet subbands forms the so-called RH signature. To justify the usefulness of the RH signature, we discuss and investigate some of its statistical properties. These properties would clarify the sufficiency of the signature to characterize the wavelet subband information. In addition, we shall also present an efficient RH signature extraction algorithm based on the coefficient-counting technique, which helps to speed up the overall classification system performance. We apply the RH signature to texture classification using the well-known databases. Experimental results show that our proposed RH signature in conjunction with the use of symmetrized Kullback-Leibler divergence gives a satisfactory classification performance compared with the current state-of-the-art methods.
25 May 2017 in Computing in Science & Engineering
CRChristopher N. Richardson
Christopher N. Richardson
Containers are an emerging technology that holds promise for improving productivity and code portability in scientific computing. The authors examine Linux container technology for the distribution of a nontrivial scientific computing software stack and its execution on a spectrum of platforms from laptop computers through high-performance computing systems. For Python code run on large parallel computers, the runtime is reduced inside a container due to faster library imports. The software distribution approach and data that the authors present will help developers and users decide on whether container technology is appropriate for them. The article also provides guidance for vendors of HPC systems that rely on proprietary libraries for performance on what they can do to make containers work seamlessly and without performance penalty.
27 October 2016 in Mathematics of Computation
The scalar and vector Laplacians are basic operators in physics and engineering. In applications, they frequently show up perturbed by lower-order terms. The effect of such perturbations on mixed finite element methods in the scalar case is well understood, but that in the vector case is not. In this paper, we first show that, surprisingly, for certain elements there is degradation of the convergence rates with certain lower-order terms even when both the solution and the data are smooth. We then give a systematic analysis of lower-order terms in mixed methods by extending the Finite Element Exterior Calculus (FEEC) framework, which contains the scalar, vector Laplacian, and many other elliptic operators as special cases. We prove that stable mixed discretization remains stable with lower-order terms for sufficiently fine discretization. Moreover, we derive sharp improved error estimates for each individual variable. In particular, this yields new results for the vector Laplacian problem which are useful in applications such as electromagnetism and acoustics modeling. Further, our results imply many previous results for the scalar problem and thus unify them all under the FEEC framework.