Improving Statistical Machine Translation Quality Using Differential Evolution

Dugonik, Jani; Bošković, Borko; Brest, Janez; Maučec, Mirjam Sepesy

doi:10.15388/informatica.2019.222

Cited by 7 publications

(4 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To improve the SMT systems' translation quality, model weights were optimized by the DE algorithm. In our previous research [59], we showed the competitive performance of the DE algorithm in comparison with MERT, MIRA, and PRO optimizers, which are commonly used in SMT optimization. The hyperparameters used to train and optimize the SMT systems are shown in Table 4.…”

Section: Experimental Settings For Models' Training and Classificationmentioning

confidence: 96%

See 1 more Smart Citation

Reduction of Neural Machine Translation Failures by Incorporating Statistical Machine Translation

et al. 2023

Self Cite

View full text Add to dashboard Cite

This paper proposes a hybrid machine translation (HMT) system that improves the quality of neural machine translation (NMT) by incorporating statistical machine translation (SMT). Therefore, two NMT systems and two SMT systems were built for the Slovenian–English language pair, each for translation in one direction. We used a multilingual language model to embed the source sentence and translations into the same vector space. From each vector, we extracted features based on the distances and similarities calculated between the source sentence and the NMT translation, and between the source sentence and the SMT translation. To select the best possible translation, we used several well-known classifiers to predict which translation system generated a better translation of the source sentence. The proposed method of combining SMT and NMT in the hybrid system is novel. Our framework is language-independent and can be applied to other languages supported by the multilingual language model. Our experiment involved empirical applications. We compared the performance of the classifiers, and the results demonstrate that our proposed HMT system achieved notable improvements in the BLEU score, with an increase of 1.5 points and 10.9 points for both translation directions, respectively.

show abstract

Section: Experimental Settings For Models' Training and Classificationmentioning

confidence: 96%

“…[57], the authors recommend using a maximum of 1000 sentences for optimization, and in Refs. [58,59], 500-700 sentences were used for optimization. We used the second part of the development set to train the classifiers and augmented the data to obtain 90,000 sentences.…”

Section: Corpora and Toolsmentioning

confidence: 99%

Reduction of Neural Machine Translation Failures by Incorporating Statistical Machine Translation

et al. 2023

Self Cite

View full text Add to dashboard Cite

show abstract

“…DE is a simple yet powerful genetic-like numerical optimizer that was proposed by Storn and Price (1997) and has become widely used (Dugonik et al, 2019;Price et al, 2006). It maintains a user-defined number (NP) of randomly-generated candidate solutions (individuals) and progressively alters them to find better ones.…”

Section: Differential Evolution (De)mentioning

confidence: 99%

A Comparative Study of Stochastic Optimizers for Fitting Neuron Models. Application to the Cerebellar Granule Cell

Cruz¹,

Marín²,

Redondo³

et al. 2021

Informatica

View full text Add to dashboard Cite

This work compares different algorithms to replace the genetic optimizer used in a recent methodology for creating realistic and computationally efficient neuron models. That method focuses on single-neuron processing and has been applied to cerebellar granule cells. It relies on the adaptive-exponential integrate-and-fire (AdEx) model, which must be adjusted with experimental data. The alternatives considered are: i) a memetic extension of the original genetic method, ii) Differential Evolution, iii) Teaching-Learning-Based Optimization, and iv) a local optimizer within a multi-start procedure. All of them ultimately outperform the original method, and the last two do it in all the scenarios considered.

show abstract

“…The most frequently used algorithms in the literature are genetic algorithm, simulated annealing, variable neighbourhood search, tabu search, etc. (see Gomes et al, 2014;Reisi-Nafchi and Moslehi, 2015;Kurdi, 2015;Zhang and Wong, 2016;Martin et al, 2016;Akbari and Rashidi, 2016;Niroomand et al, 2016;Quintana et al, 2017;Hsieh, 2017;Hu et al, 2016;Ghadiri Nejad and Banar, 2018;Misevičius et al, 2018;Vizvári et al, 2018;Dugonik et al, 2019;Ullah et al, 2020;Hassanpour, 2020;Aliya et al, 2020;Fernández et al, 2020;Hussain and Khan, 2020).…”

Section: Introductionmentioning

confidence: 99%

Tabu Search Based Hybrid Meta-Heuristic Approaches for Schedule-Based Production Cost Minimization Problem for the Case of Cable Manufacturing Systems

Daneshdoost¹,

Hajiaghaei–Keshteli²,

Şahin³

et al. 2022

Informatica

View full text Add to dashboard Cite

This paper models and solves the scheduling problem of cable manufacturing industries that minimizes the total production cost, including processing, setup, and storing costs. Two hybrid meta-heuristics, which combine simulated annealing and variable neighbourhood search algorithms with tabu search algorithm, are proposed. Applying some case-based theorems and rules, a special initial solution with optimal setup cost is obtained for the algorithms. The computational experiments, including parameter tuning and final experiments over the benchmarks obtained from a real cable manufacturing factory, show superiority of the combination of tabu search and simulated annealing comparing to the other proposed hybrid and classical meta-heuristics.

show abstract

Improving Statistical Machine Translation Quality Using Differential Evolution

Cited by 7 publications

References 20 publications

Reduction of Neural Machine Translation Failures by Incorporating Statistical Machine Translation

Reduction of Neural Machine Translation Failures by Incorporating Statistical Machine Translation

A Comparative Study of Stochastic Optimizers for Fitting Neuron Models. Application to the Cerebellar Granule Cell

Tabu Search Based Hybrid Meta-Heuristic Approaches for Schedule-Based Production Cost Minimization Problem for the Case of Cable Manufacturing Systems

Contact Info

Product

Resources

About