Abstract

Data-And model-driven computer simulations are increasingly critical in many application domains. These simulations may track 10s or 100s of parameters, affected by complex inter-dependent dynamic processes. Moreover, decision makers usually need to run large simulation ensembles, containing 1000s of simulations. In this paper, we rely on a tensor-based framework to represent and analyze patterns in large simulation ensemble data sets to obtain a high-level understanding of the dynamic processes implied by a given ensemble of simulations.We, further, note that the inherent sparsity of the simulation ensembles (relative to the space of potential simulations one can run) constitutes a significant problem in discovering these underlying patterns. To address this challenge, we propose a partition-stitch sampling scheme, which divides the parameter space into subspaces to collect several lower modal ensembles, and complement this with a novel Multi-Task Tensor Decomposition (M2TD), technique which helps effectively and efficiently stitch these subensembles back. Experiments showed that, for a given budget of simulations, the proposed structured sampling scheme leads to significantly better overall accuracy relative to traditional sampling approaches, even when the user does not have a perfect information to help guide the structured partitioning process.

Original languageEnglish (US)
Title of host publicationProceedings - IEEE 34th International Conference on Data Engineering, ICDE 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1156-1167
Number of pages12
ISBN (Electronic)9781538655207
DOIs
StatePublished - Oct 24 2018
Event34th IEEE International Conference on Data Engineering, ICDE 2018 - Paris, France
Duration: Apr 16 2018Apr 19 2018

Other

Other34th IEEE International Conference on Data Engineering, ICDE 2018
CountryFrance
CityParis
Period4/16/184/19/18

Fingerprint

Tensors
Sampling
Decomposition
Computer simulation
Simulation
Experiments

Keywords

  • Simulation
  • Tensor
  • Tensor Decomposition

ASJC Scopus subject areas

  • Information Systems
  • Information Systems and Management
  • Hardware and Architecture

Cite this

Li, X., Candan, K., & Sapino, M. L. (2018). M2td: Multi-task tensor decomposition for sparse ensemble simulations. In Proceedings - IEEE 34th International Conference on Data Engineering, ICDE 2018 (pp. 1156-1167). [8509327] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICDE.2018.00106

M2td : Multi-task tensor decomposition for sparse ensemble simulations. / Li, Xinsheng; Candan, Kasim; Sapino, Maria Luisa.

Proceedings - IEEE 34th International Conference on Data Engineering, ICDE 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 1156-1167 8509327.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Li, X, Candan, K & Sapino, ML 2018, M2td: Multi-task tensor decomposition for sparse ensemble simulations. in Proceedings - IEEE 34th International Conference on Data Engineering, ICDE 2018., 8509327, Institute of Electrical and Electronics Engineers Inc., pp. 1156-1167, 34th IEEE International Conference on Data Engineering, ICDE 2018, Paris, France, 4/16/18. https://doi.org/10.1109/ICDE.2018.00106
Li X, Candan K, Sapino ML. M2td: Multi-task tensor decomposition for sparse ensemble simulations. In Proceedings - IEEE 34th International Conference on Data Engineering, ICDE 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 1156-1167. 8509327 https://doi.org/10.1109/ICDE.2018.00106
Li, Xinsheng ; Candan, Kasim ; Sapino, Maria Luisa. / M2td : Multi-task tensor decomposition for sparse ensemble simulations. Proceedings - IEEE 34th International Conference on Data Engineering, ICDE 2018. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 1156-1167
@inproceedings{0c4a33467893440984bc6bd4ef89b073,
title = "M2td: Multi-task tensor decomposition for sparse ensemble simulations",
abstract = "Data-And model-driven computer simulations are increasingly critical in many application domains. These simulations may track 10s or 100s of parameters, affected by complex inter-dependent dynamic processes. Moreover, decision makers usually need to run large simulation ensembles, containing 1000s of simulations. In this paper, we rely on a tensor-based framework to represent and analyze patterns in large simulation ensemble data sets to obtain a high-level understanding of the dynamic processes implied by a given ensemble of simulations.We, further, note that the inherent sparsity of the simulation ensembles (relative to the space of potential simulations one can run) constitutes a significant problem in discovering these underlying patterns. To address this challenge, we propose a partition-stitch sampling scheme, which divides the parameter space into subspaces to collect several lower modal ensembles, and complement this with a novel Multi-Task Tensor Decomposition (M2TD), technique which helps effectively and efficiently stitch these subensembles back. Experiments showed that, for a given budget of simulations, the proposed structured sampling scheme leads to significantly better overall accuracy relative to traditional sampling approaches, even when the user does not have a perfect information to help guide the structured partitioning process.",
keywords = "Simulation, Tensor, Tensor Decomposition",
author = "Xinsheng Li and Kasim Candan and Sapino, {Maria Luisa}",
year = "2018",
month = "10",
day = "24",
doi = "10.1109/ICDE.2018.00106",
language = "English (US)",
pages = "1156--1167",
booktitle = "Proceedings - IEEE 34th International Conference on Data Engineering, ICDE 2018",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - M2td

T2 - Multi-task tensor decomposition for sparse ensemble simulations

AU - Li, Xinsheng

AU - Candan, Kasim

AU - Sapino, Maria Luisa

PY - 2018/10/24

Y1 - 2018/10/24

N2 - Data-And model-driven computer simulations are increasingly critical in many application domains. These simulations may track 10s or 100s of parameters, affected by complex inter-dependent dynamic processes. Moreover, decision makers usually need to run large simulation ensembles, containing 1000s of simulations. In this paper, we rely on a tensor-based framework to represent and analyze patterns in large simulation ensemble data sets to obtain a high-level understanding of the dynamic processes implied by a given ensemble of simulations.We, further, note that the inherent sparsity of the simulation ensembles (relative to the space of potential simulations one can run) constitutes a significant problem in discovering these underlying patterns. To address this challenge, we propose a partition-stitch sampling scheme, which divides the parameter space into subspaces to collect several lower modal ensembles, and complement this with a novel Multi-Task Tensor Decomposition (M2TD), technique which helps effectively and efficiently stitch these subensembles back. Experiments showed that, for a given budget of simulations, the proposed structured sampling scheme leads to significantly better overall accuracy relative to traditional sampling approaches, even when the user does not have a perfect information to help guide the structured partitioning process.

AB - Data-And model-driven computer simulations are increasingly critical in many application domains. These simulations may track 10s or 100s of parameters, affected by complex inter-dependent dynamic processes. Moreover, decision makers usually need to run large simulation ensembles, containing 1000s of simulations. In this paper, we rely on a tensor-based framework to represent and analyze patterns in large simulation ensemble data sets to obtain a high-level understanding of the dynamic processes implied by a given ensemble of simulations.We, further, note that the inherent sparsity of the simulation ensembles (relative to the space of potential simulations one can run) constitutes a significant problem in discovering these underlying patterns. To address this challenge, we propose a partition-stitch sampling scheme, which divides the parameter space into subspaces to collect several lower modal ensembles, and complement this with a novel Multi-Task Tensor Decomposition (M2TD), technique which helps effectively and efficiently stitch these subensembles back. Experiments showed that, for a given budget of simulations, the proposed structured sampling scheme leads to significantly better overall accuracy relative to traditional sampling approaches, even when the user does not have a perfect information to help guide the structured partitioning process.

KW - Simulation

KW - Tensor

KW - Tensor Decomposition

UR - http://www.scopus.com/inward/record.url?scp=85057089216&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85057089216&partnerID=8YFLogxK

U2 - 10.1109/ICDE.2018.00106

DO - 10.1109/ICDE.2018.00106

M3 - Conference contribution

SP - 1156

EP - 1167

BT - Proceedings - IEEE 34th International Conference on Data Engineering, ICDE 2018

PB - Institute of Electrical and Electronics Engineers Inc.

ER -