检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Appicharla, Ramakrishna Gain, Baban Pal, Santanu Ekbal, Asif Bhattacharyya, Pushpak Department of Computer Science and Engineering Indian Institute of Technology Patna India Wipro AI Lab45 London United Kingdom School of AI and Data Science Indian Institute of Technology Jodhpur India Department of Computer Science and Engineering Indian Institute of Technology Bombay India

In document-level neural machine translation (DocNMT), multi-encoder approaches are common in encoding context and source sentences. Recent studies (Li et al., 2020) have shown that the context encoder generates noise and makes the model robust to the choice of context. This paper further investigates this observation by explicitly modelling context encoding through multi-task learning (MTL) to make the model sensitive to the choice of context. We conduct experiments on cascade MTL architecture, which consists of one encoder and two decoders. Generation of the source from the context is considered an auxiliary task, and generation of the target from the source is the main task. We experimented with German–English language pairs on News, TED, and Europarl corpora. Evaluation results show that the proposed MTL approach performs better than concatenation-based and multi-encoder DocNMT models in low-resource settings and is sensitive to the choice of context. However, we observe that the MTL models are failing to generate the source from the context. These observations align with the previous studies, and this might suggest that the available document-level parallel corpora are not context-aware, and a robust sentence-level model can outperform the context-aware models. © 2024, CC0.

关键词： Neural machine translation

来源：评论

学校读者我要写书评

暂无评论

Pixel-Level Polyp Segmentation Network Based on Parallel Feature Enhancement and Attention Mechanism

SSRN

引用

SSRN 2024年

作者： Xu, Haiping Wang, Jie Li, Zuoyong Cheng, Xuesong Teng, Shenghua College of Computer and Data Science Minjiang University Fuzhou China School of Information and Communication Engineering University of Electronic Science and Technology of China Chengdu China College of Electronic and Information Engineering Shandong University of Science and Technology Qingdao China Fujian Key Laboratory of Medical Big Data Engineering Fujian Provincial Hospital Fuzhou China

Polyp segmentation, a crucial tool in medical diagnostics, effectively locates and segments polyps, thereby aiding clinicians in enhancing both detection efficiency and accuracy. To address the persistent challenges of imprecise polyp localization and ambiguous boundary segmentation, we propose a pixel-level polyp segmentation network based on a parallel feature enhancement and attention mechanism (PFEA-Net). The proposed PFEA-Net follows an enconder-decoder architecture. Given that polyps vary greatly in size and shape, PFEA-Net first utilizes the HarDNet68 as the encoder to extract multi-level feature representations. Then, a parallel feature enhancement module is designed to enrich the feature map of each level by integrating semantic information derived from higher-level features. Meanwhile, a parallel partial decoder module is used to introduce global contextual information. Finally, the reverse attention module refines the boundary features. PFEA-Net effectively extracts high-level representations to enhance feature depiction, capturing both detailed and semantic information in the images. Rigorous experiments conducted on five publicly available polyp datasets using six evaluation metrics conclusively demonstrate that the proposed network exhibits superior robustness and achieves more precise segmentation. Compared to other state-of-the-art algorithms, PFEA-Net distinguishes itself with exceptional performance. © 2024, The Authors. All rights reserved.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

AI-Powered Personalized Deep Learning Framework for Early Detection of Glioblastoma Using Multi-Modal MRI data

AI-Powered Personalized Deep Learning Framework for Early De...

引用

data science, Agents & Artificial Intelligence (ICDSAAI), International Conference on

作者： Sunitha T Veeramalai Sankaradass V.K Manindra Manish Department of Artificial Intelligence and Data Science Saveetha Engineering College Chennai India Dept. of Computer Science and Engineering Chennai Institute of Technology Chennai India Bachelor of Medicine and Bachelor of Science Dhanalakshmi Srinivasan Institute of Medical Sciences and Hospital Perambalur India

ISBN: (数字)9798331537555

ISBN: (纸本)9798331537562

The prognosis of glioblastoma, which stands as the most malignant variant of a brain tumour, is clearly unfavourable due to the late diagnosis and the severe heterogeneity it is known to possess. Standard diagnostic techniques nowadays that employ MRI data quite often have a low-level performance in terms of diagnosing glioblastoma during its early stages owing to the diverse manner in which the tumours may grow in various patients. In light of this background, we present a new deep learning framework suitable for multi-modal MRI data and aimed at improving glioblastoma prognosis at an early stage. The framework utilizes convolutional neural networks (CNNs) to obtain features from T1, T2, and FLAIR MRI images, and combines this with an AI-guided individualized diagnostic system. This model utilizes relevant clinical data such as age and genetic data in a more individualized manner to meet patient’s needs during diagnosis. One more novel feature of our approach is the ensemble learning that combines CNNs with RNNs to detect minor tumour deposits more effectively over time. The proposed framework delivered an AUC of 0.94 on large glioblastoma MRI databases, thus progressing the early-stage diagnosis of the tumour type which was the aim. This advanced personalized method not only allows for superior diagnosis but also improves on the treatment plans that are available for such patients. Further development will concern clinical verification as well as increasing the capacity of the framework.

关键词： Deep learning Accuracy Sensitivity Magnetic resonance imaging Genetics Feature extraction History Prognostics and health management Standards Tumors

来源：评论

学校读者我要写书评

暂无评论

How to Select Pre-Trained Code Models for Reuse? A Learning Perspective

How to Select Pre-Trained Code Models for Reuse? A Learning ...

引用

IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER)

作者： Zhangqian Bi Yao Wan Zhaoyang Chu Yufei Hu Junyi Zhang Hongyu Zhang Guandong Xu Hai Jin National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab Wuhan China School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China School of Big Data and Software Engineering Chongqing University Chongqing China School of Computer Science University of Technology Sydney Sydney Australia

ISBN: (数字)9798331535100

ISBN: (纸本)9798331535117

Pre-training a language model and then fine-tuning it has shown to be an efficient and effective technique for a wide range of code intelligence tasks, such as code generation, code summarization, and vulnerability detection. However, pre-training language models on a large-scale code corpus is compu-tationally expensive. Fortunately, many off-the-shelf Pre-trained Code Models (PCMs), such as CodeBERT, CodeT5, CodeGen, and Code Llama, have been released publicly. These models acquire general code understanding and generation capability during pre-training, which enhances their performance on downstream code intelligence tasks. With an increasing number of these public pre-trained models, selecting the most suitable one to reuse for a specific task is essential. In this paper, we systematically investigate the reusability of PCMs. We first explore three intuitive model selection methods that select by size, training data, or brute-force fine-tuning. Experimental results show that these straightforward techniques either perform poorly or suffer high costs. Motivated by these findings, we explore learning-based model selection strategies that utilize pre-trained models without altering their parameters. Specifically, we train proxy models to gauge the performance of pre-trained models, and measure the distribution deviation between a model's latent features and the task's labels, using their closeness as an indicator of model transferability. We conduct experiments on 100 widely-used open-source PCMs for code intelligence tasks, with sizes ranging from 42.5 million to 3 billion parameters. The results demonstrate that learning-based selection methods reduce selection time to 100 seconds, compared to 2,700 hours with brute-force fine-tuning, with less than 6% performance degradation across related tasks.

关键词： Phase change materials Analytical models Adaptation models Codes Costs Computational modeling Training data Machine learning Software Software development management

来源：评论

学校读者我要写书评

暂无评论

SPECTRAL NORMALIZED-CUT GRAPH PARTITIONING WITH FAIRNESS CONSTRAINTS

arXiv

引用

arXiv 2023年

作者： Li, Jia Wang, Yanhao Merchant, Arpit School of Data Science and Engineering East China Normal University Shanghai China Department of Computer Science University of Helsinki Helsinki Finland

Normalized-cut graph partitioning aims to divide the set of nodes in a graph into k disjoint clusters to minimize the fraction of the total edges between any cluster and all other clusters. In this paper, we consider a fair variant of the partitioning problem wherein nodes are characterized by a categorical sensitive attribute (e.g., gender or race) indicating membership to different demographic groups. Our goal is to ensure that each group is approximately proportionally represented in each cluster while minimizing the normalized cut value. To resolve this problem, we propose a two-phase spectral algorithm called FNM. In the first phase, we add an augmented Lagrangian term based on our fairness criteria to the objective function for obtaining a fairer spectral node embedding. Then, in the second phase, we design a rounding scheme to produce k clusters from the fair embedding that effectively trades off fairness and partition quality. Through comprehensive experiments on nine benchmark datasets, we demonstrate the superior performance of FNM compared with three baseline methods. © 2023, CC BY-SA.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

An Empirical Examination of Balancing Strategy for Counterfactual Estimation on Time Series 41

An Empirical Examination of Balancing Strategy for Counterfa...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Huang, Qiang Meng, Chuizheng Cao, Defu Huang, Biwei Chang, Yi Liu, Yan School of Artificial Intelligence Jilin University Jilin Changchun China International Center of Future Science Jilin University Jilin Changchun China Department of Computer Science University of Southern California California Los Angeles United States Halicioğlu Data Science Institute University of California San Diego San DiegoCA United States Engineering Research Center of Knowledge-Driven Human-Machine Intelligence MOE Jilin Changchun China

Counterfactual estimation from observations represents a critical endeavor in numerous application fields, such as healthcare and finance, with the primary challenge being the mitigation of treatment bias. The balancing strategy aimed at reducing covariate disparities between different treatment groups serves as a universal solution. However, when it comes to the time series data, the effectiveness of balancing strategies remains an open question, with a thorough analysis of the robustness and applicability of balancing strategies still lacking. This paper revisits counterfactual estimation in the temporal setting and provides a brief overview of recent advancements in balancing strategies. More importantly, we conduct a critical empirical examination for the effectiveness of the balancing strategies within the realm of temporal counterfactual estimation in various settings on multiple datasets. Our findings could be of significant interest to researchers and practitioners and call for a reexamination of the balancing strategy in time series settings. Copyright 2024 by the author(s)

关键词： Time series analysis

来源：评论

学校读者我要写书评

暂无评论

Passenger Waiting Time Prediction Based on Density Spatiotemporal Clustering and Time Series Feature Processing 6

Passenger Waiting Time Prediction Based on Density Spatiotem...

引用

6th International Conference on Electronic engineering and Informatics, EEI 2024

作者： Huang, Ruixi Xia, Dawen Geng, Jian Zhang, Wenyong Hu, Yang Li, Yantao Li, Huaqing College of Data Science and Information Engineering Guizhou Minzu University Guiyang550025 China Engineering Research Center of Micro-nano and Intelligent Manufacturing Ministry of Education Kaili University Kaili556011 China College of Big Data Engineering College of Microelectronics and Artificial Intelligence Kaili University Kaili556011 China Party School of the Anshun Municipal Committee of C.P.C Anshun561000 China Guizhou Traffic Technician and Transportation College Department of Automotive Engineering Guiyang550008 China College of Computer Science Chongqing University Chongqing400044 China College of Electronic and Information Engineering Southwest University Chongqing400715 China

ISBN: (纸本)9798350353594

Accurate passenger waiting time prediction addresses the taxi supply-demand imbalance in Intelligent Transport Systems (ITS). To capture the dependencies of passenger waiting time on urban taxi distribution, we propose a Density Spatiotemporal Clustering and Time Series Feature Processing (DSC-TSFP) method that discovers potential feature relationships of taxi GPS trajectory data. Specifically, a Multilevel Spatiotemporal Density-Based Spatial Clustering of Applications with Noise (MST-DBSCAN) is proposed to cluster taxi GPS trajectory data to obtain waiting time series with spatiotemporal correlations. Furthermore, a time series processing module based on the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise Normalization (CEEMDANN) is designed to enhance prediction accuracy by handling non-stationarity and numerical discrepancy, which solves the residual white noise problem in Intrinsic Modal Functions (IMFs) of EEMD and CEEMD. Next, a waiting time prediction module oriented to a Convolutional Long Short-Term Memory Network (ConvLSTM) is incorporated to compensate for the defects of spatially localized dependencies that LSTM cannot capture. Finally, a CEEMDANN-ConvLSTM model is constructed to predict the passenger waiting time by combining the time series processing and waiting time prediction module. Based on real-world taxi GPS trajectory data, experimental results demonstrate that DSC-TSFP outperforms comparable methods regarding prediction accuracy. © 2024 IEEE.

关键词： Global positioning system

来源：评论

学校读者我要写书评

暂无评论

Prior-Guided Parallel Residual Bi-Fusion Network in USV Obstacle Detection

Prior-Guided Parallel Residual Bi-Fusion Network in USV Obst...

引用

IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW)

作者： Chih-Chung Hsu Sophia Yang Xiu-Yu Hou Yu-An Jhang Institute of Data Science National Cheng Kung University Interdisciplinary Program of Electrical Engineering and Computer Science National Tsing Hua University

In this paper, we propose a novel Prior-Guided Parallel Residual Bi-Fusion Feature Pyramid Network (PPRB-FPN) for accurate obstacle detection in unmanned surface vehicle (USV) sailing. Our method tackles the challenge of detecting small objects, which are prone to information vanishing. To the end, we leverage the PRB-FPN for small object detection and YOLOv7 as a single-stage object detector to effectively identify obstacles. Our experimental results on the Obstacle Detection Challenge dataset at the 1st Workshop on Maritime computer Vision (MaCVi) demonstrate that our method outperforms both Mask R-CNN (mrcnn) and YOLOv7, achieving an F_avg score of 0.514.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Automatic liver-vessel examination from CT slice using Kapur's thresholding and watershed algorithm 2

Automatic liver-vessel examination from CT slice using Kapur...

引用

2nd International Conference on Machine Learning and data engineering, ICMLDE 2023

作者： Kadry, Seifedine Abualigah, Laith González Crespo, Rubén Verdú, Elena Damasevicius, Robertas Singh, Vijendra Rajinikanth, Venkatesan Department of Applied Data Science Noroff University College Kristiansand Norway Ajman University Ajman346 United Arab Emirates Department of Electrical and Computer Engineering Lebanese American University Byblos Lebanon Computer Science Department Al Al-Bayt University Mafraq25113 Jordan Hourani Center for Applied Scientific Research Al-Ahliyya Amman University Amman19328 Jordan MEU Research Unit Middle East University Amman11831 Jordan School of Computer Sciences Universiti Sains Malaysia Pulau Pinang11800 Malaysia School of Engineering and Technology Sunway University Malaysia Petaling Jaya27500 Malaysia Higher School of Engineering and Technology Universidad Internacional de la Rioja Logroño26006 Spain Department of Applied Informatics Vytautas Magnus University Kaunas53361 Lithuania School of Computer Science University of Petroleum and Energy Studies Dehradun248007 India Department of Computer Science and Engineering Division of Research and Innovation Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Chennai602105 India

computerized medical image examination (CMIE) plays a significant role in modern hospitals to achieve the necessary tasks, like segmentation and classification. By segmenting an image, we can extract a particular section for examination. A two-dimensional computed tomography (CT) slice was used for liver-vessel examination (LiVE). A simple automatic technique for supporting LiVE is being developed in this research. A CT slice is collected, a 3D to 2D conversion is done, (ii) Kapur's tri-level thresholding and Hummingbird-Optimizer is used to enhance the CT slice, (iii) the watershed algorithm (WA) is used to extract the vessel, and (iv) the WA is compared and verified against the segmentation methods chosen. WA provides better segmentation results than other methods because it is an automatic approach. Using the chosen image database, the proposed technique achieves an overall segmentation accuracy of >97%. Other segmentation problems can be used in the future to verify the merit of this scheme. © 2024 Elsevier B.V.. All rights reserved.

关键词： Medical imaging

来源：评论

学校读者我要写书评

暂无评论

Learning Universal Multi-level Market Irrationality Factors to Improve Stock Return Forecasting

arXiv

引用

arXiv 2025年

作者： Yang, Chen Wang, Jingyuan Jiang, Xiaohan Wu, Junjie School of Computer Science and Engineering Beihang University Beijing China School of Computer Science and Engineering School of Economics and Management Beihang University Beijing China MIIT Key Laboratory of Data Intelligence and Management School of Economics and Management Beihang University Beijing China

Recent years have witnessed the perfect encounter of deep learning and quantitative trading has achieved great success in stock investment. Numerous deep learning-based models have been developed for forecasting stock returns, leveraging the powerful representation capabilities of neural networks to identify patterns and factors influencing stock prices. These models can effectively capture general patterns in the market, such as stock price trends, volume-price relationships, and time variations. However, the impact of special irrationality factors – such as market sentiment, speculative behavior, market manipulation, and psychological biases – has not been fully considered in existing deep stock forecasting models due to their relative abstraction as well as lack of explicit labels and data description. To fill this gap, we propose UMI, a Universal multi-level Market Irrationality factor model to enhance stock return forecasting. The UMI model learns factors that can reflect irrational behaviors in market from both individual stock and overall market levels. For the stock-level, UMI construct an estimated rational price for each stock, which is cointegrated with the stock’s actual price. The discrepancy between the actual and the rational prices serves as a factor to indicate stock-level irrational events. Additionally, we define market-level irrational behaviors as anomalous synchronous fluctuations of stocks within a market. Using two self-supervised representation learning tasks, i.e., sub-market comparative learning and market synchronism prediction, the UMI model incorporates market-level irrationalities into a market representation vector, which is then used as the market-level irrationality factor. We also developed a forecasting model that captures both temporal and relational dependencies among stocks, accommodating the UMI factors. Extensive experiments on U.S. and Chinese stock markets with competitive baselines demonstrate our model’s effectiveness and

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：