The paper presents our submission to the WMT2021 Shared Task on Quality Estimation (QE). We participate in sentence-level predictions of human judgments and post-editing effort. We propose a glass-box approach based on attention weights extracted fro
m machine translation systems. In contrast to the previous works, we directly explore attention weight matrices without replacing them with general metrics (like entropy). We show that some of our models can be trained with a small amount of a high-cost labelled data. In the absence of training data our approach still demonstrates a moderate linear correlation, when trained with synthetic data.
This paper describes our submission for the shared task on Unsupervised MT and Very Low Resource Supervised MT at WMT 2021. We submitted systems for two language pairs: German ↔ Upper Sorbian (de ↔ hsb) and German-Lower Sorbian (de ↔ dsb). For de ↔ h
sb, we pretrain our system using MASS (Masked Sequence to Sequence) objective and then finetune using iterative back-translation. Final finetunng is performed using the parallel data provided for translation objective. For de ↔ dsb, no parallel data is provided in the task, we use final de ↔ hsb model as initialization of the de ↔ dsb model and train it further using iterative back-translation, using the same vocabulary as used in the de ↔ hsb model.
This paper presents the NICT Kyoto submission for the WMT'21 Quality Estimation (QE) Critical Error Detection shared task (Task 3). Our approach relies mainly on QE model pretraining for which we used 11 language pairs, three sentence-level and three
word-level translation quality metrics. Starting from an XLM-R checkpoint, we perform continued training by modifying the learning objective, switching from masked language modeling to QE oriented signals, before finetuning and ensembling the models. Results obtained on the test set in terms of correlation coefficient and F-score show that automatic metrics and synthetic data perform well for pretraining, with our submissions ranked first for two out of four language pairs. A deeper look at the impact of each metric on the downstream task indicates higher performance for token oriented metrics, while an ablation study emphasizes the usefulness of conducting both self-supervised and QE pretraining.
Dialogue policy optimisation via reinforcement learning requires a large number of training interactions, which makes learning with real users time consuming and expensive. Many set-ups therefore rely on a user simulator instead of humans. These user
simulators have their own problems. While hand-coded, rule-based user simulators have been shown to be sufficient in small, simple domains, for complex domains the number of rules quickly becomes intractable. State-of-the-art data-driven user simulators, on the other hand, are still domain-dependent. This means that adaptation to each new domain requires redesigning and retraining. In this work, we propose a domain-independent transformer-based user simulator (TUS). The structure of TUS is not tied to a specific domain, enabling domain generalization and the learning of cross-domain user behaviour from data. We compare TUS with the state-of-the-art using automatic as well as human evaluations. TUS can compete with rule-based user simulators on pre-defined domains and is able to generalize to unseen domains in a zero-shot fashion.
Abstract Large-scale pretraining and task-specific fine- tuning is now the standard methodology for many tasks in computer vision and natural language processing. Recently, a multitude of methods have been proposed for pretraining vision and language
BERTs to tackle challenges at the intersection of these two key areas of AI. These models can be categorized into either single-stream or dual-stream encoders. We study the differences between these two categories, and show how they can be unified under a single theoretical framework. We then conduct controlled experiments to discern the empirical differences between five vision and language BERTs. Our experiments show that training data and hyperparameters are responsible for most of the differences between the reported results, but they also reveal that the embedding layer plays a crucial role in these massive models.
In this study, we present a state-of-art model; we call SYRIA, to simulate the activity of ventricular myocardial cell as an example of simulating a human cell, in which we use the latest mathematical models of cardiac cell. We rely on O'Hara (O'Hara
, et al., 2011) for modeling electrical activity, ions hemostasis, and contracting. Our presented model takes into consideration the role of potassium channels KATP, chloride channels, volume regulation channels based on the Kyoto model (A.Takeuchi, 2006), PH regulation channels based on Leem model (Leem, et al., 1999), and the improvement of the values of some variables based on the results of modern experiments, especially concentrations of ions within the mitochondrial and cytoplasm, the values of calcium buffers in the SR, values of the conductance of membrane channels, and concentrations of metabolites in the mitochondria. The previous models have been linked to a mitochondrial model based on Kembro (Kembro, et al., 2013). The SYRIA model is based on the integration and improvement of the best known models in a hierarchical structure that facilitates understanding, monitoring and reuse, we also present models for testing drugs and some external influences. The programming process is done using blocks of M-file and S-function in Simulink.
By comparing the results obtained from the simulation with the laboratory results, we observe that computer simulations give results within the normal physiological range
.
This research demonstrated a way to study the thermal properties
of raw and recycled High Density Polyethylene HDPE which is
used to make oil containers. Also, the research has dealt with the
effect of recycling on blow molding process used to manufacture
these containers.
This paper addresses failure-prone
production line with limited buffers. Each machine is subject to
either time-dependent failures or operation dependent failures.
In this research a performance of TurboExpander and Joule-
Thomson valve will be compared in a proposed system developed
in order to recover flare gas in oil fields outstations which not
connected to any gas plant and burns continually the entire
associated gases in the flare, and reuses the mentioned gases in
central process facilities as a fuel in gas turbines which use diesel
(as they have dual system gas-diesel), while the associated gases in
related outstations are burned.
In this paper, we propose modifications to TCP by
introducing adaptive delay and loss response (TCP-ADaLR) to
mitigate the adverse effects of satellite link characteristics.
