检索结果-内蒙古大学图书馆

Intelligent Cybernetics Technology & Applications (ICICyTA), International Conference on

作者： Bintang Kevin Hizkia Samosir Sani Muhamad Isa Computer Science Department Binus Graduate Program – Master of Computer Science Binus University Jakarta Indonesia

ISBN: (数字)9798331506490

ISBN: (纸本)9798331506506

This study investigates the performance of Vision Transformer (ViT) variants—the Shifted Window Transformers (SWIN), Distillation with No Labels (DINO), and Data-efficient Image Transformers (DeIT)—in image captioning tasks using the Flickr8K dataset. While ViT architectures have shown promise in image classification, their effectiveness for image captioning, particularly with smaller datasets, remains unclear. The models' performance was evaluated using BLEU metrics, while training efficiency was analyzed through Pareto front analysis of computational time and accuracy. Among the tested variants, SWIN Transformers demonstrated superior performance (BLEU-1: 64.4, BLEU-2: 33.9, BLEU-3: 17.1, BLEU-4: 8.4), followed by DINO (BLEU-1: 63.1, BLEU-2: 32.7, BLEU-3: 16.4, BLEU-4: 7.5), while DeIT showed the weakest performance (BLEU-1: 61.6, BLEU-2: 31.1, BLEU-3: 14.7, BLEU-4: 6.5). SWIN Transformers achieved the shortest training time at 3 minutes 31 seconds per epoch, making it the most efficient model among ViT variants based on Pareto front analysis. While ViT variants achieved competitive BLEU-1 scores comparable to previous top models, they struggled with generating coherent, longer sentences, as evidenced by suboptimal BLEU-4 scores. These findings provide empirical evidence of how the lack of inductive bias in transformer architectures affects their ability to capture complex scene relationships, despite their strong feature detection capabilities, contributing to the understanding of transformer models' limitations in vision-language tasks, especially with limited data.

关键词： Training Measurement Analytical models Feature detection Computational modeling computer architecture Transformers Data models Computational efficiency Image classification

来源：评论

学校读者我要写书评

暂无评论

Leveraging EEG Signals and Machine Learning for Schizophrenia Classification 6

Leveraging EEG Signals and Machine Learning for Schizophreni...

引用

6th International Conference on Computing and Informatics, ICCI 2024

作者： Elfarsy, Abdelrahman El-Metwally, Sara Medical Informatics Program Faculty of Computers and Information Mansoura University Mansoura Egypt Computer Science Department Faculty of Computers and Information Mansoura University Mansoura Egypt

ISBN: (纸本)9798350373875

Schizophrenia is a neurological disorder known for its potential to disrupt brain function and cause erratic behavior. Timely diagnosis and intervention are crucial for improving patient outcomes. This paper conducts a comprehensive comparative study of three machine learning models: Logistic Regression (LR), Support Vector Machine (SVM), and Random Forest (RF) to classify EEG signals associated with schizophrenia. The dataset utilized in this study comprises EEG data obtained from 28 individuals, data preprocessing techniques, including artifact removal, filtering, and normalization, were applied to enhance data quality. A set of informative statistical features was extracted from the EEG signals to capture relevant information. The three machine learning models are trained and evaluated using various performance metrics, including accuracy, precision, recall, F1-score, and Area Under the Curve (AUC). Random Forest achieved the highest accuracy (96%), while SVM demonstrated strong precision and recall (95%). These findings highlight the potential of machine learning in aiding early schizophrenia diagnosis through EEG signal analysis. © 2024 IEEE.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Automated classification of big X-ray diffraction data using deep learning models

引用

npj Computational Materials 2023年第1期9卷 169-180页

作者： Jerardo E.Salgado Samuel Lerman Zhaotong Du Chenliang Xu Niaz Abdolrahim Materials Science program University of RochesterRochesterNew York 14627USA Department of Computer Science University of RochesterRochesterNew York 14627USA Department of Mechanical Engineering University of RochesterRochesterNew York 14627USA Laboratory for Laser Energetics University of RochesterRochesterNew York 14627USA

In current in situ X-ray diffraction(XRD)techniques,data generation surpasses human analytical capabilities,potentially leading to the loss of *** techniques require human intervention,and lack the performance and adaptability required for material *** the critical need for high-throughput automated XRD pattern analysis,we present a generalized deep learning model to classify a diverse set of materials’crystal systems and space *** our approach,we generate training data with a holistic representation of patterns that emerge from varying experimental conditions and crystal *** also employ an expedited learning technique to refine our model’s expertise to experimental *** addition,we optimize model architecture to elicit classification based on Bragg’s Law and use evaluation data to interpret our model’s *** evaluate our models using experimental data,materials unseen in training,and altered cubic crystals,where we observe state-of-the-art performance and even greater advances in space group classification.

关键词： Automate classify cubic

来源：评论

学校读者我要写书评

暂无评论

The Country's Implementation and Adoption of Standardized Health Terminologies to Promote Interoperability: A Systematic Literature Review 8

The Country's Implementation and Adoption of Standardized He...

引用

8th International Conference on Information Management and Technology, ICIMTech 2023

作者： Sitompul, Taufiq Hendric Spits Warnars, Harco Leslie Meyliana Hidayanto, Achmad Nizar Prabowo, Harjanto Bina Nusantara University Binus Graduate Program - Doctor of Computer Science Computer Science Department Jakarta11480 Indonesia School of Information Systems Bina Nusantara University Information Systems Department Jakarta11480 Indonesia Universitas Indonesia Faculty of Computer Science Depok16424 Indonesia Binus Business School Undergraduate Program Bina Nusantara University Management Department Jakarta11480 Indonesia

ISBN: (纸本)9798350326093

Modern healthcare systems demand comprehensive information systems but face obstacles during adoption. Organizational and structural complexity, especially decentralized systems, challenges the integrated management and sharing of data across information systems. Interoperability is critical to address these issues effectively in heterogeneous and distributed environments. However, different medical terminology hinders interoperability within health information systems. To address this, a systematic literature review examined various medical terminology standards for national interoperability. From 790 articles, 12 were selected according to PRISMA guidelines with inclusion and exclusion criteria. Standard health terminology is critical to achieving interoperability, and many countries have adopted it, although some have required adaptation to accommodate local contexts. Data models, governance, and maintenance of standards are equally important in achieving national interoperability. In addition, standardized health terminology promotes consistency and uniformity in terminology, which is beneficial for decision-making by stakeholders in the health system. © 2023 IEEE.

关键词： Interoperability

来源：评论

学校读者我要写书评

暂无评论

Gender-Based Bimodal Analysis of Speech and Text in Non-Contextual Emotion Recognition

Gender-Based Bimodal Analysis of Speech and Text in Non-Cont...

引用

2024 Beyond Technology Summit on Informatics International Conference, BTS-I2C 2024

作者： Angkasa, Abby Rasya Zahra, Amalia Fung, Wai-Keung Binus Graduate Program Computer Science Department Jakarta11480 Indonesia Cardiff Metropolitan University Cardiff School of Technologies Engineering Department Cardiff United Kingdom

ISBN: (纸本)9798331508579

Emotion recognition can help human-computer interactions by enabling systems to respond empathetically and adapt to users' emotional conditions. This capability improves user experience, supporting the development of a more intuitive and emotionally responsive communication system. This study analyzes a bimodal approach based on gender (male and female) to recognize emotions without contextual information in dialogue analysis. Utilizing the Multimodal EmoryNLP dataset extracted from the TV series Friends with acted speech, we focused on four primary emotions: Angry, Neutral, Joy, and Scared. The model used in this study for text classification is RoBERTa, and wav2vec 2.0 is used for audio feature extraction with the Bi-LSTM model for classification. The experiment results using weighted F1-score reveal that data augmentation enhanced the performance of analyzing the original dataset from 0.46% to 0.52% and the male dataset from 0.43% to 0.51 %. In comparison, the female dataset remained consistent at 0.46%. The weighted F1-score and Unweighted Averaged Recall (UAR) from the male dataset are higher, 51 % and 48%, respectively, than those from the female dataset, 46% and 47%, respectively. Gender-based analysis indicated that male and female datasets exhibited distinct performance patterns, highlighting variations in emotional expression and recognition between genders. These findings underscore the effectiveness of multimodal strategies in emotion recognition and suggest that gender-specific factors play a significant role in enhancing classification performance. While these results highlight performance trends, further validation through repeated trials and statistical analyses could provide stronger generalizations and insights into gender-based differences. © 2024 IEEE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

Optimizing Random Forest Classification Using Chi-Square and SMOTE-ENN on Student Drop-Out Data 8

Optimizing Random Forest Classification Using Chi-Square and...

引用

8th International Conference on Informatics and Computing, ICIC 2023

作者： Andri Yunis, Roni Tanti Universitas Mikroskil Undergraduate Program of Computer Science Department Medan Indonesia Universitas Mikroskil Undergraduate Program of Information Systems Department Medan Indonesia Universitas Mikroskil Undergraduate Program of Information Technology Department Medan Indonesia

ISBN: (纸本)9798350342604

Dropout is a particular concern for countries striving to increase human capital. Various attempts have been made by universities to minimize the number of dropouts. Machine learning has also developed various predictive models to determine the likelihood of students dropping out. However, there is a challenge in dropout data, specifically the problem of class imbalance, where the number of students who drop out (minority class) is significantly less than those who do not drop out (majority class). This imbalance can reduce the model's ability to classify students at risk of dropping out. This study proposes classification optimization using the Random Forest algorithm to handle class imbalances in student dropout data. To overcome class imbalance, the Synthetic Minority Oversampling Technique (SMOTE) and Edited Nearest Neighbors (ENN) techniques are used. Additionally, the attribute selection method is also applied to enhance the predictive results. The test results demonstrate that the combination of implementing feature selection with Chi-Square, followed by class imbalance handling with SMOTE-ENN, provided the most optimal predictive performance for identifying the status of both dropouts and graduates. © 2023 IEEE.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

A Machine Learning-Based Approach to Calibrate Low-Cost Particulate Matter Sensors 12

A Machine Learning-Based Approach to Calibrate Low-Cost Part...

引用

12th Brazilian Symposium on Computing Systems Engineering, SBESC 2022

作者： Pastorio, Andre F. Spanhol, Fabio A. Martins, Leila D. De Camargo, Edson T. Graduate Program in Computer Science Cascavel Brazil Toledo Brazil Londrina Brazil

ISBN: (数字)9781665474252

ISBN: (纸本)9781665474252

Low-cost particulate matter (LC-PM) sensors have been studied around the world as a viable alternative to expensive reference stations for monitoring air quality. However, LC-PM sensors require periodic calibration, since their data are often inaccurate and subject to uncertainty. Sensors calibration can be performed through machine learning methods where the sensor is placed in a real environment subject to the local environmental conditions of the place and its measurement compared to a reference equipment. This work evaluates different machine learning methods in five different models of LC-PM sensors, aiming to select the most appropriate sensor and a calibration method to be used in a low-cost air quality station in the context of smart cities. © 2022 IEEE.

关键词： Calibration

来源：评论

学校读者我要写书评

暂无评论

Cybersecurity Detection Using Neural Networks for Chemical Industry 1

Cybersecurity Detection Using Neural Networks for Chemical I...

引用

1st International Conference on Technology, Engineering, and Computing Applications, ICTECA 2023

作者： Pratama, Ilham Fitrah Eka Mantoro, Teddy Wijayanti, Rika Persadha, Pratama Dahlian Postgraduate Program of Medical Intelligence National Intelligence College Bogor Indonesia Sampoerna University Department of Computer Science Jakarta Indonesia

ISBN: (纸本)9798350360554

Cybersecurity threats affect all sectors of the economy, especially the chemical industry. The chemical industry is a critical infrastructure vulnerable to cyberattacks. In addition to financial and reputational losses, they can also cause serious accidents (such as loss of control of hazardous materials) with serious consequences for workers, society, and the environment. Industrial control panel systems are typically embedded in a variety of physical devices. The safety of control systems is crucial for ensuring the secure functioning of industrial chemical processes. Nevertheless, the susceptibility of control systems to cyber-attacks has intensified due to the extensive deployment of actuators, sensor measurements, and wireless communications. These attacks have the potential to propagate swiftly and result in severe industrial accidents. This study was conducted as part of a systematic literature review. The purpose of this study is to analyze cybersecurity threats to the chemical industry. Research shows that although cyber attacks are highly vulnerable to the chemical industry, they can be detected through neural network-based integration (NN) without having to know the process model when large quantities of historical data are available and easy to apply. It has been shown that it can be done. It is hoped that the policy recommendations obtained from this research will provide more effective guidelines for protecting the chemical industry from the risks of cyber threats. © 2023 IEEE.

关键词： Chemical industry

来源：评论

学校读者我要写书评

暂无评论

The Role of Biosafety and Biosecurity in Biotechnology to Prevent Bioterrorism Threats 1

The Role of Biosafety and Biosecurity in Biotechnology to Pr...

引用

1st International Conference on Technology, Engineering, and Computing Applications, ICTECA 2023

作者： Subariyah, Riski Mantoro, Teddy Ratmono, Barito Mulyo Postgraduate Program of Medical Intelligence National Intelligence College Bogor Indonesia Sampoerna University Department of Computer Science Jakarta Indonesia

ISBN: (纸本)9798350360554

'Society 5.0' is characterized as a high degree of integration of cyberspace and physical space capable of promoting economic growth and addressing social problems in a modern information society based on cutting-edge technology advancements for data processing and knowledge creation. Recent developments in the field of biotechnology have been demonstrated by gene editing techniques, especially CRISPR-Cas9, such as basic research, applied biotechnology, and biomedical research Most biotechnology is essentially 'dual use' because its development can result in both profitable and nefarious goals, all depending on the perspective and motivation of the perpetrator. The purpose of this research is a better understanding of biosecurity, biosecurity, and bioterrorism threats, so as to provide a foundation for policymakers, researchers, and health professionals to improve global health security so as to realize effective biodefense systems. The research was conducted by library research, namely data collection with literature studies. By implementing biosafety and biosecurity laboratory measures in accordance with established guidelines and guidelines can help protect people, animals, the environment, and national security. Medical intelligence is vital to monitoring and assessing the threat of bioterrorism through the process of data collection, which is then analyzed, and data dissemination. The medical intelligence community can help protect public safety and prevent bioterrorism by understanding and implementing strong biosafety and biosecurity practices in laboratories. Bioterrorism attacks that can occur at any time must be anticipated by medical intelligence by conducting research and assessment of bioterrorism threats, analyzing information so that it can be passed on to the government to take appropriate policies to be able to detect and monitor the spread of infectious diseases quickly, and developing strategies to prevent the spread of disease. Combining health s

关键词： Bioterrorism

来源：评论

学校读者我要写书评

暂无评论

The Utilization of Fuzzy Logic and Bollinger Bands to Enhance Trading Decision-Making During the Bitcoin Halving Phase 9th

The Utilization of Fuzzy Logic and Bollinger Bands to Enhanc...

引用

9th International Conference on computer science and Computational Intelligence, ICCSCI 2024

作者： Darmawan, Oscar Adam Heryadi, Yaya Lukas Wulandhari, Lili Ayu Sonata, Ilvico Computer Science Department Binus Graduate Program Bina Nusantara University Jakarta11480 Indonesia Universitas Katolik Indonesia Atma Jaya Jakarta12930 Indonesia

The volatile behavior of Bitcoin's price, especially during its halving periods, poses considerable obstacles for forecasting and decision-making in cryptocurrency trading. This paper presents a novel method that combines application fuzzy logic with Bollinger Bands to improve trading decision-making in times of high market volatility. This study conducted an experiment utilizing three fuzzy logic controllers and Bollinger Bands (BB) to determine the strength of buy, hold, and sell signals. This dataset includes the initial and final prices that are used to calculate the BB. The raw and computed values serve as the precise input parameters for the Fuzzy Inference System (FIS). The membership functions were categorized into four levels: very low, low, high, and very high, based on the input default settings utilized by traders. Rulesets were created using fuzzy logic to produce signals that indicate the level of strength of a trading advice. This study evaluate the effectiveness of this hybrid method in comparison to the traditional utilization of the Bollinger Band only indicator and Moving Average Convergence Divergence (MACD) indicator, which is widely favored by traders to identify possible market fluctuations. This methodology involves creating a trading simulation that is based on past Bitcoin halving events. The objective is to assess the efficacy of these strategies in managing heightened volatility. The application of fuzzy logic with the Bollinger Bands model yielded a success rate of 92.47% while analyzing 93 daily data points from the previous Bitcoin halving event on May 11, 2020. © 2024 The Authors.

关键词： Costs

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：