Probabilistic Data-Driven Sampling via Multi-Criteria Importance Analysis

Biswas, Ayan; Dutta, Soumya; Lawrence, Earl; Patchett, John; Calhoun, Jon C.; Ahrens, James

doi:10.1109/tvcg.2020.3006426

Cited by 16 publications

(21 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Different solutions have been proposed, including kernel methods [30] and binning [12]. Although the binning strategy is scalable with respect to the number of data points, the quality of the estimation may be dependent on the number of bins used [31]. Kernel methods offer a more continuous description of the PDF but typically scale poorly with the number of data points, as local density value requires knowledge of the relative locations of all data points.…”

Section: Probability Map Estimationmentioning

confidence: 99%

“…They first estimate the probability density function (PDF) of the scalar values and then use it to downselect data points. The method was later extended to include the scalar gradients to ensure visual smoothness [31]. This technique requires binning the phase-space to construct the PDF, which poses two main issues.…”

Section: Introductionmentioning

confidence: 99%

“…This technique requires binning the phase-space to construct the PDF, which poses two main issues. First, the choice of the number of bins can affect the quality of the results [31]. Second, binning high-dimensional spaces can become intractable in terms of computational intensity and memory, which makes extension to uniform sampling in higher scalar dimensions nontrivial [32].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Uniform-in-Phase-Space Data Selection with Iterative Normalizing Flows

Hassanaly¹,

Perry²,

Müeller³

et al. 2021

Preprint

View full text Add to dashboard Cite

Figure 1: Graphical illustration of the data selection method. Starting from a large dataset (top left scatter plot), a reduced uniform-in-phase-space dataset is constructed (right scatter plot). The reduction method relies on the probability map which is iteratively computed (Step 1 and Step 1 correction). An acceptance probability constructed with the probability map is used to accept or reject each data point (Step 2).

show abstract

Section: Probability Map Estimationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Uniform-in-Phase-Space Data Selection with Iterative Normalizing Flows

Hassanaly¹,

Perry²,

Müeller³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Nguyen and Song [17] incorporated centrality-driven clustering information during random sampling. Using the ideas of entropy maximization, Biswas et al [4,5] recently proposed in situ data-driven sampling schemes that preserve important data features along with their gradient properties. For scattered datasets, Rapp et al [20] proposed a blue noise preserving sampling method to identify representative subset of points.…”

Section: Related Workmentioning

confidence: 99%

“…Instead of storing the highresolution data sets, the corresponding sampled data is stored and subsequently used for various post-hoc analyses. Different flavors of sampling algorithms exist in literature [4,5,7,20] which can selectively sample different data properties. However, most of these algorithms primarily target univariate data.…”

Section: Introductionmentioning

confidence: 99%

Relationship-aware Multivariate Sampling Strategy for Scientific Simulation Data

Hazarika

Biswas

Wolfram

et al. 2020

2020 IEEE Visualization Conference (VIS)

Self Cite

View full text Add to dashboard Cite

With the increasing computational power of current supercomputers, the size of data produced by scientific simulations is rapidly growing. To reduce the storage footprint and facilitate scalable post-hoc analyses of such scientific data sets, various data reduction/summarization methods have been proposed over the years. Different flavors of sampling algorithms exist to sample the high-resolution scientific data, while preserving important data properties required for subsequent analyses. However, most of these sampling algorithms are designed for univariate data and cater to post-hoc analyses of single variables. In this work, we propose a multivariate sampling strategy which preserves the original variable relationships and enables different multivariate analyses directly on the sampled data. Our proposed strategy utilizes principal component analysis to capture the variance of multivariate data and can be built on top of any existing state-of-the-art sampling algorithms for single variables. In addition, we also propose variants of different data partitioning schemes (regular and irregular) to efficiently model the local multivariate relationships. Using two real-world multivariate data sets, we demonstrate the efficacy of our proposed multivariate sampling strategy with respect to its data reduction capabilities as well as the ease of performing efficient post-hoc multivariate analyses.

show abstract