检索结果-内蒙古大学图书馆

6th International Seminar on Research of Information Technology and Intelligent Systems, ISRITI 2023

作者： Ali, Fadel Arymurthy, Aniati Murni Prasojo, Radityo Eko Universitas Indonesia Faculty of Computer Science Depok Indonesia Pitik Digital Indonesia Department of AI & Data South Tangerang Indonesia

ISBN: (纸本)9798350358346

Speech Emotion Recognition (SER) is a pivotal area in Human-computer Interaction (HCI) with numerous applications. Traditional SER models rely on supervised learning but face challenges due to limited labeled data and the subjective nature of emotions. Self-supervised learning (SSL) offers an alternative by leveraging unlabeled audio data. WavLM, a large-scale SSL audio model, has shown promise in various speech processing tasks. This paper investigates the fine-Tuning of WavLM for SER, analyzing the impact of different segment of layers on performance. We conducted experiments on the IEMOCAP and RAVDESS datasets, comparing various fine-Tuned WavLM models with Wav2Vec 2.0 and HuBERT as SSL baselines. Results reveal that WavLM outperforms SSL base-lines, with all-layer fine-Tuned WavLM achieving state-of-The-Art results. Interestingly, fine-Tuning top layers significantly enhances performance, suggesting their role in encoding paralinguistic information. However, fine-Tuning all layers remains superior. These findings shed light on optimizing SSL audio models for SER and highlight WavLM's potential in emotion recognition. © 2023 IEEE.

关键词： Supervised learning

来源：评论

学校读者我要写书评

暂无评论

The role of data governance in a high-level approach of data migration to open data

引用

Business Information Review 2024年第3期41卷 94-103页

作者： Yulfitri, Alivia Sensuse, Dana Indra Sumirat Hidayat, Deden Suryono, Ryan Randy Kautsarina Purwaningsih, Erisva Hakiki Prabuwono, Anton Satria Faculty of Computer Science Universitas Indonesia Depok Indonesia Faculty of Computer Science Universitas Esa Unggul West Jakarta Indonesia Research Center for Data and Information Sciences National Research and Innovation Agency (BRIN) Central Jakarta and Bandung Indonesia Faculty of Engineering and Computer Science Universitas Teknokrat Indonesia Lampung Indonesia Information and Communications Technology Human Resources Development Agency Ministry of Communication and Information Technology of the Republic of Indonesia Jakarta Central Jakarta Indonesia Faculty of Computing and Information Technology in Rabigh King Abdulaziz University Jeddah Saudi Arabia

data quality issues are still a challenge in data migration. Existing methods mainly focus on technical issues. However, the problem of data quality is not only a technical problem, but the main one is the non-technical aspect. This paper aims to identify the role of data governance in the data migration process. The qualitative research method uses two primary references, namely the high-level data migration approach from Kazimir et al. and the DAMA-DMBOK framework. Both are highly compatible with data migration processes focusing on data quality and referencing governance roles. The case study was conducted in a central government working on data migration to the new One data application. The results show that eight data governance roles are involved in 20 migration processes (65%). Non-technical activities related to data quality involve more than half of the data migration process. This paper proposes the role of data governance for non-technical aspects of the migration process. This paper presents a case study of the high-level data migration process and how data governance comes into play. This research can help the government and practitioners overcome problems that often occur in data migration, especially related to data quality. © The Author(s) 2024.

关键词： DAMA-DMBOK framework data governance data migration Indonesia’s one data Kazimir et al.’s approach open data

来源：评论

学校读者我要写书评

暂无评论

Hypernym Information and Sentiment Bias Probing in Distributed data Representation 23

Hypernym Information and Sentiment Bias Probing in Distribut...

引用

15th International Conference on Machine Learning and Computing, ICMLC 2023

作者： Acquaye, Frank Lawrence Nii Adoquaye Latipov, Insan-Aleksandr Kertész-Farkas, Attila Department of Data Analysis and Artificial Intelligence Laboratory on AI for Computational Biology Faculty of Computer Science HSE University Moscow Russia

ISBN: (纸本)9781450398411

Neural word embedding vectors have been exhaustively investigated by probing tasks, whether they contain semantic and syn- tactic information. Perhaps the most popular task is a test on gender relation "king - man + woman ≈ queen", other probings include tests on singular/plural relation (apple∼apples), analogy (good:better∼rough:X), purity of the clusters of word embeddings (categorization). Here, we propose two novel probing tasks to evaluate the compositionality of word embeddings. First, we propose probing tests to evaluate whether word embeddings contain structural information about their hypernyms;more specifically, whether it is possible to predict the hypernym (e.g. color) of a word (e.g. red) using word embedding. Our experimental results show that a simple logistic regression can predict well the hypernyms of words;therefore, making the word embeddings linearly separable with respect to their hypernyms. Second, we also provide a methodology to investigate whether word embeddings encode undesired sentiment. For instance, we show that the word embedding of American or Euro- pean are surrounded by more positive words than the embeddings around the word "Chinese". Thus, we conclude that word embed- dings also can capture sentiment, leading to undesired machine bias in downstream applications. © 2023 ACM.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

Classification of Sleep-Related Disorders Using Machine Learning Techniques 25

Classification of Sleep-Related Disorders Using Machine Lear...

引用

25th International Arab Conference on Information Technology, ACIT 2024

作者： Nasir, Ola Alazaidah, Raed Al-Safarini, Maram Y. Alamleh, Amneh Zaid, Ali Abu Mansur, Hasan Elrashidi, Ali Alzoubi, Haneen Bani-Bakr, Alaa Zarqa University Faculty of Information Technology Department of Computer Science Zarqa Jordan Zarqa University Faculty of Information Technology Department of Data Science and Ai Zarqa Jordan Zarqa University Faculty of Information Technology Department of Cyber Security Zarqa Jordan American University of Madaba Faculty of Business Department of Accounting Madaba Jordan University of Business and Technology Department of Electrical Engineering Jeddah21432 Saudi Arabia Private University Faculty of It Applied Science Department of Computer Science Amman Jordan

ISBN: (纸本)9798331540012

Several people suffer from irregular sleep patterns, a problem that negatively affects their daily lives. It presents a multitude of challenges, including daytime distress, disturbances in sleep-wake patterns, feelings of unease, and a variety of other complications. Diabetes, obesity, and psychological disorders are among the diseases that may be responsible for this sleep disorder. Therefore, the primary aim of this research is to employ the robust capabilities of machine learning in forecasting sleep disorders. This research classifies three diseases: diabetes, obesity, and psychological disorders, using thirty-two different classifiers belonging to five learning strategies. The research compares and evaluates their predictive performance on four datasets. The comparative analysis includes four evaluation metrics: Accuracy, Precision, ROC Area, and MCC. The research results indicate that the best classifier is SGD, followed by LMT, while NBM is the weakest classifier. In addition, the research results revealed that the best strategy is Functions, followed by Rules, with Bayes being the least effective strategy. While conducting the experiment and analyzing the data, it is observed that the best feature selection method that yielded the most favorable outcome is the CorrelationAttributeEval. Furthermore, it has been found that the classification algorithm known as SGD achieved the highest Accuracy and Precision in categorizing the data. This indicates that the combination of CorrelationAttributeEval as the feature selection approach and SGD as the classification algorithm proved to be the most successful and efficient strategy for this research © 2024 IEEE.

关键词： Sleep research

来源：评论

学校读者我要写书评

暂无评论

Enhanced Pediatric Pneumonia Diagnosis with Chest X-Ray using Deep Attention Mechanism 11

Enhanced Pediatric Pneumonia Diagnosis with Chest X-Ray usin...

引用

11th IEEE/ACM International Conference on Big data Computing, Applications and Technologies, BDCAT 2024

作者： Zaman, Muhammad Fatima, Tahseen Hameed, Sana Haider, Shahzeb Akhunzada, Adnan Azeem, Muhammad University of Lahore Faculty of Computer Science Sargodha Pakistan University of Saint Thomas School of Engineering Saint PaulMN United States University of Doha for Science and Technology Department of Data and Cybersecurity Doha24449 Qatar

ISBN: (纸本)9798350367300

Development of deep learning technologies has resulted in a major revolution in the field of medical diagnostics in recent years. The purpose of this research is to examine the usage of Convolutional Neural Networks (CNNs) in combination with an integrated attention mechanism known as CoATNet to improve the accuracy and responsiveness of paediatric pneumonia diagnosis using chest X-ray analysis. The major goal of this research project is to build and verify a powerful deep learning model capable of detecting mild pneumonia symptoms in children's chest X-rays. CoATNet's innovative attention approach allows it to provide more weight to areas deemed to be extremely important. This is in addition to its ability to recognise major patterns in images. Furthermore, the paper considers the issues posed by class imbalances in medical imaging datasets and undertakes an in-depth assessment of how effectively CoATNet works across a large dataset of paediatric chest X-rays. CoATNet's remarkable performance was shown by its ability to achieve a Micro Average of 98.2%, a Weighted Average of 98.7%, and an Accuracy of 99.7%. These measures show that the model can distinguish between instances of pneumonia and those do not have the illness, and that it is resilient when faced with class imbalances. F1 Score of 98.9% reflects a well-balanced trade-off between Precision 98.7% and Recall 98.2%. The integration of deep learning technologies with attention processes in CoATNet has resulted in a big step forward in the development of diagnostic models that are more visible and useful. © 2024 IEEE.

关键词： Pediatrics

来源：评论

学校读者我要写书评

暂无评论

Dual Dynamic Consistency Regularization for Semi-Supervised Domain Adaptation

引用

IEEE Access 2024年 12卷 36267-36279页

作者： Ngo, Ba Hung Lam, Ba Thinh Nguyen, Thanh Huy Dinh, Quang Vinh Choi, Tae Jong Chonnam National University Graduate School of Data Science Gwangju61186 Korea Republic of University of Science Faculty of Information Technology Ho Chi Minh City700000 Viet Nam National Cheng Kung University Department of Biomedical Engineering Tainan70101 Taiwan Vietnamese-German University School of Electrical Engineering and Computer Science Ho Chi Minh City72000 Viet Nam

The Vision Transformer (ViT) model serves as a powerful model to capture and comprehend global information, particularly when trained on extensive datasets. Conversely, the Convolutional Neural Network (CNN) model is beneficial to training with small datasets for retaining essential local information. Inspired by these properties of ViT and CNN models, we introduce a hybrid framework that smoothly increases the cross-domain generalization in Semi-supervised Domain Adaptation (SSDA). To achieve this goal, we first train the ViT model on abundant labeled source data, while the CNN model is trained on a few labeled target samples. Then, these models dynamically exchange their knowledge for potential generalization to unlabeled target data via the proposed method, named Dual Dynamic Consistency Regularization (D2CR). Specifically, the ViT model provides its pseudo labels to update the global perspective for the CNN model. Similarly, the CNN model offers pseudo labels to complement the local perspective for the ViT model. The previous methods use a fixed threshold algorithm for the pseudo-labeling process. However, we utilize the dynamic threshold strategy to create pseudo labels for the bi-directional consistency regularization learning between the ViT and CNN models. We verify our approach across several SSDA benchmark datasets. The outstanding experimental results provide strong evidence of the effectiveness and superiority of our approach over previous state-of-the-art SSDA methods. © 2013 IEEE.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

Anomaly-Free Prior Guided Knowledge Distillation for Industrial Anomaly Detection

Anomaly-Free Prior Guided Knowledge Distillation for Industr...

引用

2024 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2024

作者： Li, Gang Chen, Tianjiao Li, Min Han, Delong Zhou, MingLe Key Laboratory of Computing Power Network and Information Security Ministry of Education Jinan China Shandong Fundamental Research Center for Computer Science Shandong Provincial Key Laboratory of Computer Networks Jinan China City University of Macau Faculty of Data Science China

ISBN: (纸本)9781665410205

In industrial manufacturing, visual anomaly detection is critical for maintaining product quality by detecting and preventing production anomalies. Anomaly detection methods based on knowledge distillation demonstrate promising performance in addressing the unpredictability and diversity of anomalies. However, they suffer from a lack of effective guidance from anomaly-free priors when handling anomalous features and underutilize multi-scale features during the segmentation scoring stage, yielding suboptimal detection results. To alleviate these issues, we propose an Anomaly-free Prior Guided knowledge distillation (APG) for industrial anomaly detection. Firstly, it filters the abnormal features by training the de-noising target network with knowledge distillation structure. Concurrently, we propose the Prior Perception Propagation Module (P3M), which extracts more efficient anomaly-free features by imposing constraints on anomalous features. Secondly, we propose the Multi-scale Prior Guided Fusion Module (MPGFM) to improve anomaly detection accuracy by utilizing anomaly-free features from the target network as priors to guide the generation and fusion of cross-scale differential features. Finally, the Global Perception Enhancement Module (GPEM) is proposed to construct an anomaly scoring network, leveraging comprehensive scene features to enhance the detection and localization performance of numerous small-target anomalies in industrial manufacturing. Extensive experiments on the MVTecAD and BTAD datasets show that the proposed method demonstrates a consistent and significant outperformance against competing methods. © 2024 IEEE.

关键词： Anomaly detection

来源：评论

学校读者我要写书评

暂无评论

A Bioinformatics Analysis and Homology 3D Modeling of Dihydroxy-orotate-dehydrogenase of Plasmodium ovale

A Bioinformatics Analysis and Homology 3D Modeling of Dihydr...

引用

2023 International Conference on Informatics, Multimedia, Cyber and Information Systems, ICIMCIS 2023

作者： Nirwantono, Rudi Assidqi, Khoirunnisa Hidayat, Alam Ahmad Pebrianto Trinugroho, Joko Pardamean, Bens Bina Nusantara University Faculty of Engineering Bioinformatics and Data Science Research Center Biotechnology Department Jakarta11480 Indonesia Bina Nusantara University Faculty of Engineering Food Biotechnology Research Center Biotechnology Department Jakarta11480 Indonesia Bina Nusantara University School of Computer Science Bioinformatics and Data Science Research Center Mathematics Departement Jakarta11480 Indonesia Bina Nusantara University Bioinformatics and Data Science Research Center Jakarta11480 Indonesia Bina Nusantara University Binus Graduate Program - Computer Science Program Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350307306

The infection of Plasmodium vivax is relatively less virulent than the deathliest Plasmodium falciparum. However, it still can lead to a fatal case and often induces recurring malaria due to dormant parasites in the liver. Thus, the research to study the drug to treat Plasmodium vivax is essential, where the enzyme dihydroorotate dehydrogenase (DHODH) has recently become a new drug target. However, the drug-enzyme interaction study has only recently been conducted in Plasmodium falciparum (pfDHODH) due to the lack of the 3D structure of enzyme DHODH from Plasmodium vivax that is crucial for the study. Therefore, this study aimed to perform the modelling study of Plasmodium vivax DHODH (PvDHODH) to create a 3D structure as a basis for drug-protein interaction study in upcoming studies. Sequence pvDHODH (Accession: SCO68359.1) was used in homology modelling using Modeller 10.2 with the crystal structure of DHODH from DHODH (Accession: 7KZY) as a template. Overall, the model generated from the homology modelling was considered a good model, and the 3D structure was close to the native state according to several parameters, including DOPE score, GA341 score, QMEAN4 value, 3D-1D score, ERRAT2 score, and ProSA score. Ramachandran plot also revealed that almost all amino acids were distributed in the desirable and allowed area (99.1%), and only 0.9% were in generously allowed regions. Superimposition of the model with the template also indicated that the model has an almost similar structure and amino acids positioning, including in the binding pocket and active site. Therefore, the model can be used for downstream analysis in drug-protein interaction studies. Nevertheless, some improvements can still be performed to upgrade the quality of the model by deleting unaligned residues on the C-terminal and realigning the residues at the doubting site. © 2023 IEEE.

关键词： Amino acids

来源：评论

学校读者我要写书评

暂无评论

Equity and Fairness of Bayesian Knowledge Tracing 15

Equity and Fairness of Bayesian Knowledge Tracing

引用

15th International Conference on Educational data Mining, EDM 2022

作者： Tschiatschek, Sebastian Knobelsdorf, Maria Singla, Adish University of Vienna Faculty of Computer Science Research Network Data Science Vienna Austria University of Vienna Faculty of Computer Science Computer Science Education Vienna Austria Max-Planck Institute for Software Systems Saarbrücken Germany

ISBN: (纸本)9781733673631

We consider the equity and fairness of curricula derived from Knowledge Tracing models. We begin by defining a unifying notion of an equitable tutoring system as a system that achieves maximum possible knowledge in minimal time for each student interacting with it. Realizing perfect equity requires tutoring systems that can provide individualized curricula per student. In particular, we investigate the design of equitable tutoring systems that derive their curricula from Knowledge Tracing models. We first show that the classical Bayesian Knowledge Tracing (BKT) model and their derived curricula can fall short of achieving equitable tutoring. To overcome this issue, we then propose a novel model, Bayesian-Bayesian Knowledge Tracing (B2KT), that naturally allows online individualization. We demonstrate that curricula derived from our model are more effective and equitable than those derived from existing models. Furthermore, we highlight that improving models with a focus on the fairness of next-step predictions can be insufficient to develop equitable tutoring systems. © 2022 Copyright is held by the author(s).

关键词： Curricula

来源：评论

学校读者我要写书评

暂无评论

MemADet: A Representative Memory Bank Approach for Industrial Image Anomaly Detection

MemADet: A Representative Memory Bank Approach for Industria...

引用

2024 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2024

作者： Li, Min He, Jinghui Ying, Zuobin Li, Gang Zhou, Mingle Key Laboratory of Computing Power Network and Information Security Ministry of Education Jinan China Shandong Fundamental Research Center for Computer Science Shandong Provincial Key Laboratory of Computer Networks Jinan China City University of Macau Faculty of Data Science China

ISBN: (纸本)9781665410205

In the field of industrial production, anomaly detection is crucial for ensuring product quality and maintaining production efficiency. With the continuous advancement of computer vision technology, it has shown tremendous potential in industrial applications. However, the scarcity of labeled anomaly samples in real-world operating environments poses significant challenges for traditional anomaly detection techniques. To address this, we propose MemADet, a novel anomaly detection model that employs an unsupervised approach and leverages a representative memory bank. It employs a dynamic decision mechanism to control the representativeness of the features stored in the memory bank, and employs a weighted anomaly score calculation mechanism to further enhance the performance of image anomaly detection. Our evaluations indicate that MemADet performs robustly in industrial image anomaly detection across three datasets, with particularly no-table detection accuracy on the MVTec AD dataset. Its efficacy is further validated by competitive results on two additional datasets, highlighting its consistent and effective performance in various settings. © 2024 IEEE.

关键词：

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：