Results 21 to 30 of about 22,610 (167)

A Comparative Study on Regularization Strategies for Embedding-based Neural Networks

, 2015
This paper aims to compare different regularization strategies to address a common phenomenon, severe overfitting, in embedding-based neural networks for NLP. We chose two widely studied neural models and tasks as our testbed. We tried several frequently
Chen, Yunchuan, Jin, Zhi, Li, Ge, Lu, Yangyang, Mou, Lili, Peng, Hao   +5 more
core   +1 more source

The regularity game: Investigating linguistic rule dynamics in a population of interacting agents [PDF]

Cognition, 2017
Rules are an efficient feature of natural languages which allow speakers to use a finite set of instructions to generate a virtually infinite set of utterances. Yet, for many regular rules, there are irregular exceptions. There has been lively debate in cognitive science about how individual learners acquire rules and exceptions; for example, how they ...
Cuskley, Christine, CASTELLANO, CLAUDIO, COLAIORI, FRANCESCA, LORETO, Vittorio, Pugliese, Martina, TRIA, FRANCESCA   +5 more
openaire   +3 more sources

Typical reconstruction limit and phase transition of maximum entropy method

Physical Review Research
We investigate the dependence of the maximum entropy method (MEM) reconstruction performance on the default model. The maximum entropy method is a reconstruction technique that utilizes prior information, referred to as the default model, to recover ...
Masaru Hitomi, Masayuki Ohzeki
doaj   +1 more source

Non-Autoregressive Machine Translation with Auxiliary Regularization

, 2019
As a new neural machine translation approach, Non-Autoregressive machine Translation (NAT) has attracted attention recently due to its high efficiency in inference.
He, Di, Liu, Tie-Yan, Qin, Tao, Tian, Fei, Wang, Yiren, Zhai, ChengXiang   +5 more
core   +1 more source

Artificial Intelligence for Bone: Theory, Methods, and Applications

Advanced Intelligent Discovery, EarlyView.
Advances in artificial intelligence (AI) offer the potential to improve bone research. The current review explores the contributions of AI to pathological study, biomarker discovery, drug design, and clinical diagnosis and prognosis of bone diseases. We envision that AI‐driven methodologies will enable identifying novel targets for drugs discovery. The
Dongfeng Yuan, Haicang Zhang, Liping Tong, Di Chen   +3 more
wiley   +1 more source

Unsupervised Neural Machine Translation with SMT as Posterior Regularization

, 2019
Without real bilingual corpus available, unsupervised Neural Machine Translation (NMT) typically requires pseudo parallel data generated with the back-translation method for the model training. However, due to weak supervision, the pseudo data inevitably
Liu, Shujie, Ma, Shuai, Ren, Shuo, Zhang, Zhirui, Zhou, Ming   +4 more
core   +1 more source

B/ordering and healthcare access for migrants with precarious status: The role of healthcare workers in counteracting restrictive policies

American Journal of Community Psychology, EarlyView.
Abstract In Canada, precarious migration is largely invisibilized. Nonetheless, b/ordering greatly affects people's realities by limiting access to social rights. In Quebec, migrants with precarious status (MPS) do not have access to healthcare, although Quebec has a “universal” healthcare coverage.
Émilie Pigeon‐Gagné, Jill Hanley, Élise Bourgeois‐Guérin, Pénélope Boudreault   +3 more
wiley   +1 more source

Revisiting Recurrent Networks for Paraphrastic Sentence Embeddings

, 2017
We consider the problem of learning general-purpose, paraphrastic sentence embeddings, revisiting the setting of Wieting et al. (2016b). While they found LSTM recurrent networks to underperform word averaging, we present several developments that ...
Gimpel, Kevin, Wieting, John
core   +1 more source

Sentence Analogies: Exploring Linguistic Relationships and Regularities in Sentence Embeddings

CoRR, 2020
While important properties of word vector representations have been studied extensively, far less is known about the properties of sentence vector representations. Word vectors are often evaluated by assessing to what degree they exhibit regularities with regard to relationships of the sort considered in word analogies. In this paper, we investigate to
Xunjie Zhu, Gerard de Melo
openaire   +2 more sources

Modified Neumann incision for implant placement in overdentures: Case series of a less invasive alternative

Clinical Advances in Periodontics, EarlyView.
Abstract Background Conventional envelope flap techniques for implant placement in edentulous mandibles are associated with greater postoperative morbidity, especially in elderly patients or those with anatomical limitations. Minimally invasive alternatives aim to preserve soft tissues and reduce complications, enhancing clinical predictability ...
Lucas Jardim da Silva, Laura Lourenço Morel, Luciana de Rezende Pinto, Otacílio Luiz Chagas‐Júnior, Jamil Awad Shibli, Fernanda Faot   +5 more
wiley   +1 more source