Elastic Net Regularization in Lorentz force evaluation

Dölker, Eva-Maria; Schmidt, Reinhard H.; Gorges, Stephan; Otterbach, Jan-Marc; Petković, Bojana; Strohmeier, Daniel; Eichardt, Roland; Brauer, Hartmut; Haueisen, Jens

doi:10.1016/j.ndteint.2018.07.002

Cited by 7 publications

(8 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The AFS shows the lowest errors for the cylindrical and the L-shaped defects (Tables I and II), where the relatively low error for the L-shaped defect might be explained by error compensation as the defect consists of two parts, one with extension in x-direction and the other with extension in y-direction. The NRMSE values increase for deeper defects, which is in line with previous work (Dölker et al, 2018b). Despite these NRMSEs, appropriate size and depth estimation of cylindrically shaped defects were conducted in (Mengelkamp et al, 2016a;Mengelkamp et al, 2016b;Dölker et al, 2018b).…”

Section: Discussionsupporting

confidence: 87%

“…The NRMSE values increase for deeper defects, which is in line with previous work (Dölker et al, 2018b). Despite these NRMSEs, appropriate size and depth estimation of cylindrically shaped defects were conducted in (Mengelkamp et al, 2016a;Mengelkamp et al, 2016b;Dölker et al, 2018b). A correct depth estimation is possible with AFS even if the estimated defect size is not correct (Dölker et al, 2018a).…”

Section: Discussionsupporting

confidence: 85%

“…The Lorentz force evaluation (LFE) is a novel non-destructive evaluation method for conducting specimen (Petkovi c et al, 2013), which allows the characterization of deep-lying defects (Dölker et al, 2018b). In LFE, a permanent magnet is moved relative to a conducting specimen.…”

Section: Introductionmentioning

confidence: 99%

“…The Lorentz force perturbation signal is referred to as defect response signal (DRS). The defect properties such as location and shape are determined by the DRS through solving an ill-posed inverse problem (Petkovi c et al, 2013;Mengelkamp et al, 2016a;Mengelkamp et al, 2016b;Dölker et al, 2018b).…”

Section: Introductionmentioning

confidence: 99%

“…The quality of the defect reconstruction depends on various factors, such as the forward method, the inverse method, the measurement data, etc. Previous studies used the approximate forward solution (AFS) (Petkovi c et al, 2013;Mengelkamp et al, 2016a;Mengelkamp et al, 2016b;Dölker et al, 2018b) and the extended area approach (EAA) (Ziolkowski, 2015;Dölker et al, 2018a) to model the DRS. Both methods are fast but limited to single, regular-shaped defects far away from the specimen edges.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Single voxel approach for Lorentz force evaluation

Dölker

Petković

Schmidt

et al. 2019

COMPEL

View full text Add to dashboard Cite

Purpose Lorentz force evaluation is a non-destructive evaluation method for conducting specimens. The movement of a specimen relative to a permanent magnet leads to Lorentz forces that are perturbed in the presence of a defect. This defect response signal (DRS) is used for defect reconstruction. To solve a linear inverse problem for defect reconstruction, an accurate and fast forward computation method is required. As existing forward methods are either too slow or too inaccurate, the purpose of this paper is to propose the single voxel approach (SVA) as a novel method. Design/methodology/approach In SVA, the DRS is computed as a superposition of DRSs from single defect voxels, which are calculated in advance, by applying the boundary element source method. This research uses a setup of an isotropic conducting specimen, a spherical permanent magnet and defects of different shapes at different depths. With the help of simulations, this study compares the SVA to the previously proposed approximate forward solution (AFS) and the extended area approach (EAA) using the normalized root mean square error (NRMSE). Simulated data using the finite element method serve as the reference solution. Findings SVA shows across all simulations NRMSE values <2.5 per cent compared to <8 per cent for EAA and <12 per cent for AFS. Originality/value The superposition principle of SVA allows for the application of linear inverse methods for defect reconstruction while providing sufficient accuracy of the forward method.

show abstract

Section: Discussionsupporting

confidence: 87%

Section: Discussionsupporting

confidence: 85%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Single voxel approach for Lorentz force evaluation

Dölker

Petković

Schmidt

et al. 2019

COMPEL

View full text Add to dashboard Cite

show abstract

Hashing-based semantic relevance attributed knowledge graph embedding enhancement for deep probabilistic recommendation

Khan

et al. 2022

Appl Intell

View full text Add to dashboard Cite

Knowledge graph embedding (KGE) is effectively exploited in providing precise and accurate recommendations from many perspectives in different application scenarios. However, such methods that utilize entire embedded Knowledge Graph (KG) without applying information-relevance regulatory constraints fail to stop the noise penetration into the underlying information. Moreover, higher computational time complexity is a CPU overhead in KG-enhanced systems and applications. The occurrence of these limitations significantly degrade the recommendation performance. Therefore, to cope with these challenges we proposed novel KGEE (Knowledge Graph Embedding Enhancement) approach of Hashing-based Semantic-relevance Attributed Graph-embedding Enhancement (H-SAGE) to model semantically-relevant higher-order entities and relations into the unique Meta-paths. For this purpose, we introduced Node Relevance-based Guided-walk (NRG) modeling technique. Further, to deal with the computational time-complexity, we converted the relevant information to the Hash-codes and proposed Deep-Probabilistic (dProb) technique to place hash-codes in the relevant hash-buckets. Again, we used dProb to generate guided function-calls to maximize the possibility of Hash-Hits in the hash-buckets. In case of Hash-Miss, we applied Locality Sensitive (LS) hashing to retrieve the required information. We performed experiments on three benchmark datasets and compared the empirical as well as the computational performance of H-SAGE with the baseline approaches. The achieved results and comparisons demonstrate that the proposed approach has outperformed the-state-of-the-art methods in the mentioned facets of evaluation.

show abstract