An LIBS quantitative analysis method for alloy steel at high temperature based on transfer learning

Yang, Jianhong; Li, Xiaomeng; Lu, Hanchao; Xu, Jinan; Li, Haixia

doi:10.1039/c8ja00069g

Cited by 28 publications

(17 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In contrast, the performances of other models are quite unstable for different datasets and different oxides, which demonstrates that our model has a superior generalization ability over the other methods. [26] 49 [25] 39…”

Section: E Experimental Results Of Other Oxidementioning

confidence: 99%

“…From Table IV and Figure 7, we can see that Hackem-LIBS improve prediction accuracy and outperforms the other two models. Therefore, the prediction accuracy of Hackem-LIBS is higher than that of the models in [25] and [26] in LIBS quantitative analysis with a better generalization ability.…”

Section: Comparison With Ensemble Learnersmentioning

confidence: 90%

“…We also compare Hackem-LIBS with the two existing ensemble learning based LIBS quantitative analysis models [25][26]. In [25], the authors collected data from lowtemperature steel samples to train their AdaBoost ensemble models and applied it to high-temperature scenarios through transfer-learning. Due to the different application scenarios between our work and theirs, it is impossible to perform a direct comparison.…”

Section: Comparison With Ensemble Learnersmentioning

confidence: 99%

“…In [26], the authors used a Bagging ensemble framework which ensembled multiple SVR models to conduct LIBS quantitative analysis, which can be directly comparable with our work. Here we also compared our model with two emsemble learners, [25] 0 It can be seen from Tab. IV that our model produces the smallest MSE and the highest R 2 value among the three models, indicating that our proposed model is robust to the noises in the data which contribute to accurate predictions.…”

Section: Comparison With Ensemble Learnersmentioning

confidence: 99%

“…In the chemical analysis community, many researchers also proposed to introduce ensemble learning to integrate multiple base learners for a betterperformance LIBS analysis. Yang J et al [25] chose SVR as the component learner, and combined Adaboost with transfer learning to analyze higher-temperature samples through training the integrated model with lowertemperature steel data. The experimental results proved that the Adaboost based ensemble model could make better predictions than the SVR model.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Hackem-LIBS: An Heterogeneous Stacking Ensemble Model for Laser-Induced Breakdown Spectroscopy Elemental Quantitative Analysis

Zeng

Gong

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Laser-Induced Breakdown Spectroscopy (LIBS) is a popular technique for elemental quantitative analysis in chemistry community, based on which, various methods are developed to determinate the concentrations of chemical samples. Despite the successful applications of the existing methods, they still struggle to obtain accurate samples analyses, due to their limited prediction capability, the complex compositions of samples and mutual interference of elements. In this paper, we propose a novel heterogeneous stacking ensemble learning model called Heterogeneous stACKing Ensemble Model LIBS (Hackem-LIBS) to achieve LIBS quantitative analysis with higher accuracy. Specifically, we propose a stacking ensemble learning framework consisting two stages. In the first stage, we train different heterogeneous component learners with multiple sub-training sets and pick out the optimal learners. In the second stage, we leverage the enhanced features predicted by the selected learners to train a stronger metalearner, which is used to make the final prediction. In addition, we combine Genetic Algorithm (GA) with Sequential Forward Selection (SFS) to reduce the redundancy of training features, which ensures more effective learning and higher computation efficiency. Extensive experiments on two public benchmarks are conducted and the results show that our approach achieves better accuracy in determinating the concentrations of elements and is practically applicable to the quantitative analysis of complex chemical samples via the LIBS technique.

show abstract

Section: E Experimental Results Of Other Oxidementioning

confidence: 99%

Section: Comparison With Ensemble Learnersmentioning

confidence: 90%