Exploiting transformation-domain sparsity for fast query in multiple-point geostatistics

Abdollahifard, Mohammad Javad; Nasiri, Behrooz

doi:10.1007/s10596-016-9612-1

“…In (Abdollahifard & Nasiri, 2017), Abdollahifard and Nasiri proposed an approach to speed up template matching for MPS methods. The key idea consists of calculating the similarity between data events and TI's patterns using a low dimension approximation of them.…”

Section: Review Of the Fast Template Matching In Transform Domain Basmentioning

confidence: 99%

“…The approaches of Abdollahifard & Nasiri (2017) and LSHSIM are quite different. The former does an exhaustive search in the set of low dimension patterns while the latter does an optimized search (RLE based) in a subset of the original patterns that are likely to be close to the data event.…”

Section: Review Of the Fast Template Matching In Transform Domain Basmentioning

confidence: 99%

“…Moreover, LSHSIM has a theoretical guarantee of finding, with high probability, a pattern that is near to the given data event. As far as we know there are no theoretical guarantees available for the search proposed by Abdollahifard & Nasiri (2017). However, a theoretical guarantee could be achieved if a random projection (Leskovec et al, 2014) is used rather than a LSHSIM: A Locality Sensitive Hashing Based Method for Multiple-Point Geostatistics 28…”

Section: Review Of the Fast Template Matching In Transform Domain Basmentioning

confidence: 99%

See 1 more Smart Citation

LSHSIM: A Locality Sensitive Hashing based method for multiple-point geostatistics

Moura

¹

,

Laber

²

,

Lopes

³

et al. 2017

Computers & Geosciences

View full text Add to dashboard Cite

Reservoir modeling is a very important task that permits the representation of a geological region of interest. Given the uncertainty involved in the process, one wants to generate a considerable number of possible scenarios so as to find those which best represent this region. Then, there is a strong demand for quickly generating each simulation. Since its inception, many methodologies have been proposed for this purpose and, in the last two decades, multiple-point geostatistics (MPS) has been the dominant one. This methodology is strongly based on the concept of training image (TI) and the use of its characteristics, which are called patterns. In this work, we propose a new MPS method that combines the application of a technique called Locality Sensitive Hashing (LSH), which permits to accelerate the search for patterns similar to a target one, with a Run-Length Encoding (RLE) compression technique that speeds up the calculation of the Hamming similarity. We have performed experiments with both categorical and continuous images which showed that LSHSIM is computationally efficient and produce good quality realizations, while achieving a reasonable space of uncertainty. In particular, for categorical data, the results suggest that LSHSIM is faster than MS-CCSIM, one of the state-of-the-art methods.

show abstract

“…In fact, we understand that (Abdollahifard & Nasiri, 2017) is much more related with CCSIM (Tahmasebi et al, 2012) since both rely on efficiently calculating convolutions. If a fast Fourier transformation (FFT) is employed, as proposed in MS-CCSIM , CCSIM executes log k operations per pattern/data event comparisons, where k is the number of pixels of the data event.…”

Section: Dctmentioning

confidence: 99%

“…If a fast Fourier transformation (FFT) is employed, as proposed in MS-CCSIM , CCSIM executes log k operations per pattern/data event comparisons, where k is the number of pixels of the data event. Thus, (Abdollahifard & Nasiri, 2017) is advantageous with respect to CCSIM, in terms of speed, if m < log k can be chosen. In terms of reproduction quality, CCSIM has the potential advantage of not losing information.…”

Section: Dctmentioning

confidence: 99%

Lshsim: A Locality Sensitive Hashing Based Method for Multiple-Point Geostatistics

MOURA¹

0

View full text Add to dashboard Cite

Reservoir modeling is a very important task that permits the representation of a geological region of interest. Given the uncertainty involved in the process, one wants to generate a considerable number of possible scenarios so as to find those which best represent this region. Then, there is a strong demand for quickly generating each simulation. Since its inception, many methodologies have been proposed for this purpose and, in the last two decades, multiple-point geostatistics (MPS) has been the dominant one. This methodology is strongly based on the concept of training image (TI) and the use of its characteristics, which are called patterns. In this work, we propose a new MPS method that combines the application of a technique called Locality Sensitive Hashing (LSH), which permits to accelerate the search for patterns similar to a target one, with a Run-Length Encoding (RLE) compression technique that speeds up the calculation of the Hamming similarity. We have performed experiments with both categorical and continuous images which showed that LSHSIM is computationally efficient and produce good quality realizations, while achieving a reasonable space of uncertainty. In particular, for categorical data, the results suggest that LSHSIM is faster than MS-CCSIM, one of the state-of-the-art methods.

show abstract

Investigating extreme scenarios with multiple-point geostatistics and variance maximization

Abdollahifard

¹

,

Mariethoz

²

,

Mohammadi

³

2019

Stoch Environ Res Risk Assess

1

0

View full text Add to dashboard Cite

Exploiting transformation-domain sparsity for fast query in multiple-point geostatistics

Cited by 6 publications

References 30 publications

LSHSIM: A Locality Sensitive Hashing based method for multiple-point geostatistics

LSHSIM: A Locality Sensitive Hashing based method for multiple-point geostatistics

Lshsim: A Locality Sensitive Hashing Based Method for Multiple-Point Geostatistics

Investigating extreme scenarios with multiple-point geostatistics and variance maximization

Contact Info

Product

Resources

About