Neural Machine Translation (NMT) performs training of a neural network employing an encoder-decoder architecture. However, the quality of the neural-based translations predominantly depends on the availability of a la...
详细信息
ISBN:
(纸本)9789897584848
Neural Machine Translation (NMT) performs training of a neural network employing an encoder-decoder architecture. However, the quality of the neural-based translations predominantly depends on the availability of a large amount of bilingual training dataset. In this paper, we explore the performance of translations predicted by attention-based NMT systems for Spanish to Persian low-resource language pairs. We analyze the errors of NMT systems that occur in the Persian language and provide an in-depth comparison of the performance of the system based on variations in sentence length and size of the training dataset. We evaluate our translation results using BLEU and human evaluation measures based on the adequacy, fluency, and overall rating.
According to the survey results of the World Health Organization, currently affected by the pressure of life and work, more than one-third of people worldwide have sleep problems of varying degrees. Among them, apnea ...
详细信息
Current efforts to improve math education often call for ambitious math instruction, which promotes student thinking and conceptual understanding. Such instruction can be difficult for many teachers because it differs...
详细信息
Unit commitment (UC) is a central market clearing process in wholesale electricity markets. With the increasing size and complexity of market-clearing models, UC problems become more and more complicated. To improve t...
详细信息
Dynamics of rotational magnetic particle alignment, an inverse process to Brownian relaxation, has been studied theoretically. It was found that arbitrary viscous torque with history term due to inertia and friction o...
详细信息
While pre-trained language models (PTLMs) have achieved noticeable success on many NLP tasks, they still struggle for tasks that require event temporal reasoning, which is essential for event-centric applications. We ...
详细信息
ISBN:
(纸本)9781955917094
While pre-trained language models (PTLMs) have achieved noticeable success on many NLP tasks, they still struggle for tasks that require event temporal reasoning, which is essential for event-centric applications. We present a continual pre-training approach that equips PTLMs with targeted knowledge about event temporal relations. We design self-supervised learning objectives to recover masked-out event and temporal indicators and to discriminate sentences from their corrupted counterparts (where event or temporal indicators got replaced). By further pre-training a PTLM with these objectives jointly, we reinforce its attention to event and temporal information, yielding enhanced capability on event temporal reasoning. This Effective CONtinual pre-training framework for Event Temporal reasoning (ECONET) improves the PTLMs' fine-tuning performances across five relation extraction and question answering tasks and achieves new or on-par state-of-the-art performances in most of our downstream tasks.(1)
Traditional Human Activity Recognition (HAR) approaches often rely on handcrafted features and incomplete feature extraction, limiting their effectiveness. To address these challenges, we propose a 2D Convolutional Ne...
详细信息
ISBN:
(数字)9798331533113
ISBN:
(纸本)9798331533120
Traditional Human Activity Recognition (HAR) approaches often rely on handcrafted features and incomplete feature extraction, limiting their effectiveness. To address these challenges, we propose a 2D Convolutional Neural Network (2D-CNN) model with an attention mechanism, incorporating Short-Time Fourier Transform (STFT) to convert raw time-series sensor data into time-frequency representations, enabling richer spatiotemporal feature extraction. The 2D-CNN is designed to capture hierarchical patterns, while the Triplet Attention mechanism learns cross-dimensional dependencies. The proposed framework is evaluated on three widely used HAR datasets, WISDM, UCI-HAR, and PAMAP2, achieving classification accuracies of 96.65%, 96.3%, and 97.32%. Experimental results demonstrate that our method outperforms existing deep learning models, delivering improved recognition accuracy and robustness across diverse activities.
The objective of the research is to develop a predictive model that can significantly enhance the detection and monitoring performance of Autism Spectrum Disorder (ASD) using four supervised learning techniques. In th...
详细信息
Predicting the relationships between drugs and targets is a crucial step in the course of drug discovery and development. computational prediction of associations between drugs and targets greatly enhances the probabi...
详细信息
ISBN:
(纸本)9783031138294;9783031138287
Predicting the relationships between drugs and targets is a crucial step in the course of drug discovery and development. computational prediction of associations between drugs and targets greatly enhances the probability of finding new interactions by reducing the cost of in vitro experiments. In this paper, a Meta-path-based Representation Learning model, namely MRLDTI, is proposed to predict unknown DTIs. Specifically, we first design a random walk strategy with a meta-path to collect the biological relations of drugs and targets. Then, the representations of drugs and targets are captured by a heterogeneous skip-gram algorithm. Finally, a machine learning classifier is employed by MRLDTI to discover novel DTIs. Experimental results indicate that MRLDTI performs better than several state-of-the-art models under ten-fold cross-validation on the gold standard dataset.
Due to the large applicability of AI systems in various applications, fairness in model predictions is extremely important to ensure that the systems work equally well for everyone. Biased feature representations migh...
详细信息
ISBN:
(纸本)9781665401913
Due to the large applicability of AI systems in various applications, fairness in model predictions is extremely important to ensure that the systems work equally well for everyone. Biased feature representations might often lead to unfair model predictions. To address the concern, in this research, a novel method, termed as Attention Aware Debiasing (AAD) method, is proposed to learn unbiased feature representations. The proposed method uses an attention mechanism to focus on the features important for the main task while suppressing the features related to the sensitive attributes. This minimizes the model's dependency on the sensitive attribute while performing the main task. Multiple experiments are performed on two publicly available datasets, MORPH and UTKFace, to showcase the effectiveness of the proposed AAD method for bias mitigation. The proposed AAD method enhances the overall model performance and reduces the disparity in model prediction across different subgroups.
暂无评论