检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Gröger, Fabian Gottfrois, Philippe Amruthalingam, Ludovic Gonzalez-Jimenez, Alvaro Lionetti, Simone Soenksen-Martinez, Luis R. Navarini, Alexander A. Pouly, Marc Department of Biomedical Engineering University of Basel Switzerland Department of Computer Science Lucerne University of Applied Sciences and Arts Switzerland Department of Dermatology University Hospital of Basel Switzerland J-Clinc for AI and Health Massachusetts Institute of Technology United States Center for Bioengineering Innovation and Design Johns Hopkins University United States

The growing demand for accurate and equitable AI models in digital dermatology faces a significant challenge: the lack of diverse, high-quality labeled data. In this work, we investigate the potential of domain-specific foundation models for dermatology in addressing this challenge. We utilize self-supervised learning (SSL) techniques to pre-train models on a dataset of over 240,000 dermatological images from public and private collections. Our study considers several SSL methods and compares the resulting foundation models against domain-agnostic models like those pre-trained on ImageNet and state-of-the-art models such as MONET across 12 downstream tasks. Unlike previous research, we emphasize the development of smaller models that are more suitable for resource-limited clinical settings, facilitating easier adaptation to a broad range of use cases. Results show that models pre-trained in this work not only outperform general-purpose models but also approach the performance of models 50 times larger on clinically relevant diagnostic tasks. To promote further research in this direction, we publicly release both the training code and the foundation models, which can benefit clinicians in dermatological applications. Copyright © 2024, The Authors. All rights reserved.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Ga2O3 nanorods synthesized by hydrothermal method for dual-functional sensing of pH value and glucose

引用

Sensors and Actuators A: Physical 2023年 363卷

作者： Chou, Hsin-Yu Yadlapalli, Bharath Kumar Yen, Chao-Chun Singh, Anoop Kumar Chiang, Jung-Lung Wuu, Dong-Sing Department of Materials Science and Engineering National Chung Hsing University Taichung40227 Taiwan Ph.D. Program Prospective Technology of Electrical Engineering and Computer Science National Chin-Yi University of Technology Taichung411030 Taiwan Department of Applied Materials and Optoelectronic Engineering National Chi Nan University Nantou54561 Taiwan

In this work, hydrothermal (HT)-synthesized gallium oxide (Ga2O3) nanorods were applied on indium tin oxide (ITO) glass substrates to create a dual-functional pH-glucose sensing device. Extended-gate field-effect transistors (EGFETs) were designed for pH and glucose sensing, and the response to pH levels ranging from 1 to 11 was tested using both chemical bath deposition (CBD) and high-pressure hydrothermal methods. Two sensor configurations, Ga2O3(CBD)/ITO/glass and Ga2O3(HT)/ITO/glass, were immersed in solution pH1 to 11 and cultured medium for hydrogen ion and glucose sensing. The results demonstrated that, for pH sensing in different samples, both IDS-VDS analysis and IDS-VGS analysis showed superior linearity and sensitivity with the HT method. For glucose detection, the HT method was selected for further experiments due to its maintained concentration-dependent relationship and selectivity towards various common sugars. When using culture medium DMEM as the sample (4.5 mg/ml glucose concentration), the accuracy of glucose estimation reached 99%. Stability and biocompatibility tests were also conducted, and the results indicated consistent performance. In summary, the HT-synthesized Ga2O3 nanorods exhibited high sensitivity, selectivity, and stability without affecting cell growth. Therefore, they hold great potential for applications not only in general blood glucose sensing but also in studying cell microenvironment metabolism. © 2023 Elsevier B.V.

关键词： Nanorods

来源：评论

学校读者我要写书评

暂无评论

A Neuroimaging ML-Based Framework for Anosmia Grading in Covid-19

A Neuroimaging ML-Based Framework for Anosmia Grading in Cov...

引用

IEEE International Symposium on Biomedical Imaging

作者： Hossam Magdy Balaha Mayada Elgendy Ahmed Alksas Mohamed Shehata Norah Saleh Alghamdi Fatma Taher Mohammed Ghazal Mahitab Ghoneim Eslam Hamed Fatma Sherif Ahmed Elgarayhi Mohammed Sallah Mohamed Abdelbadie Salem Elsharawy Kamal Harpal Sandhu Ayman El-Baz Department of Bioengineering J.B. Speed School of Engineering University of Louisville Louisville USA Applied Theoretical Physics Research Group Physics Department Faculty of Science Mansoura University Egypt Department of Computer Sciences College of Computer and Information Sciences Princess Nourah Bint Abdulrahman University Saudi Arabia The College of Technological Innovation Zayed University Dubai UAE Electrical Computer and Biomedical Engineering Depatrment Abu Dhabi University UAE Department of Radiology Faculty of Medicine Mansoura University Egypt Otolaryngology Department Faculty of Mediine Mansoura University Egypt Higher Institute of Engineering and Technology Egypt

ISBN: (数字)9798350313338

ISBN: (纸本)9798350313345

SARS-CoV-2 is the cause of the acute respiratory illness COVID-19. In order to identify whether a COVID-19 patient has normal, moderate, or severe anosmia, this study suggests a computer-assisted diagnostic system. It utilizes fluid-attenuated inversion recovery Magnetic Resonance Imaging and Diffusion Tensor Imaging to extract the olfactory nerve’s appearance, morphology, and diffusivity markers. The suggested CAD system consists of the following stages: (1) extract appearance markers, morphology markers, and diffusivity markers, (2) apply markers fusion, and (3) select the most promising classifier as the basis for the decision and the relevant performance indicators. The current work is unique in that it uses an ensemble of the trained, tuned, and improved ML classifiers to identify OB anosmia using majority voting. From the fusion tests using different folds, stacking boosted the weight accuracy, weighted precision, weighted specificity, weighted sensitivity, and weighted F1-score to 85.30%, 80%, 89.50%, 78.90%, and 78.30% respectively. Regarding the various performance criteria, the LGBM outperforms other ML classifiers.

关键词： COVID-19 Neuroimaging Sensitivity Shape Stacking Olfactory Morphology

来源：评论

学校读者我要写书评

暂无评论

Efficient or Powerful? Trade-offs Between Machine Learning and Deep Learning for Mental Illness Detection on Social Media

arXiv

引用

arXiv 2025年

作者： Ding, Zhanyi Wang, Zhongyan Zhang, Yeyubei Cao, Yuchen Liu, Yunchong Shen, Xiaorui Tian, Yexin Dai, Jianglai University of Pennsylvania School of Engineering and Applied Science United States New York University Center for Data Science United States Georgia Institute of Technology College of Computing United States University of California Berkeley Department of EECS United States Northeastern University Khoury College of Computer Science United States

Social media platforms provide valuable insights into mental health trends by capturing user-generated discussions on conditions such as depression, anxiety, and suicidal ideation. Machine learning (ML) and deep learning (DL) models have been increasingly applied to classify mental health conditions from textual data, but selecting the most effective model involves trade-offs in accuracy, interpretability, and computational efficiency. This study evaluates multiple ML models, including logistic regression, random forest, and LightGBM, alongside deep learning architectures such as ALBERT and Gated Recurrent Units (GRUs), for both binary and multi-class classification of mental health conditions. Our findings indicate that ML and DL models achieve comparable classification performance on medium-sized datasets, with ML models offering greater interpretability through variable importance scores, while DL models are more robust to complex linguistic patterns. Additionally, ML models require explicit feature engineering, whereas DL models learn hierarchical representations directly from text. Logistic regression provides the advantage of capturing both positive and negative associations between features and mental health conditions, whereas tree-based models prioritize decision-making power through split-based feature selection. This study offers empirical insights into the advantages and limitations of different modeling approaches and provides recommendations for selecting appropriate methods based on dataset size, interpretability needs, and computational constraints. © 2025, CC BY.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Modelling distillate output of a solar distiller with eco-friendly wood-based steam generation basin using machine learning model and starling murmuration optimizer

International Journal of Thermofluids

引用

International Journal of Thermofluids 2025年 27卷

作者： Al-Betar, Mohammed Azmi Muskens, Otto L. Hammoodi, Karrar A. Abd Elaziz, Mohamed Shambour, Qusai Yousef Fraihat, Salam Elsheikh, Ammar H. College of Engineering and Information Technology Ajman University Ajman United Arab Emirates Faculty of Engineering and Physical Sciences University of Southampton Southampton SO17 1BJ United Kingdom College of Engineering University of Al Maarif Al Anbar31001 Iraq Department of Mathematics Faculty of Science Zagazig University Zagazig44519 Egypt Faculty of Computer Science and Engineering Galala University Suez435611 Egypt Software Engineering Department Hourani Center for Applied Scientific Research Al-Ahliyya Amman University Amman Jordan Department of Production Engineering and Mechanical Design Faculty of Engineering Tanta University Tanta31527 Egypt Faculty of Engineering Pharos University in Alexandria Egypt

Yield augmentation of solar distiller (SD) using low-cost modifications with minimum environmental impacts is highly desirable. In this study, we developed an eco-friendly steam generation device made of carbonized Ficus sycomorus wood to boost the yield of the conventional single-basin solar distiller. Moreover, a hybrid model composed of a long short-term memory (LSTM) network and a Starling murmuration optimizer (SMO) was developed to predict the yield of the established SD. To verify the accuracy of the model, three other models were employed and the distillate yield was predicted. The predicted data of all models was evaluated using different error measures. These models were standalone LSTM and LSTM optimized by an artificial hummingbird algorithm (AHA) or manta ray foraging optimizer (MRFO). Real field data were employed during the training and testing of all models. The thermal performance of the distiller with the steam generation device was compared with that of a traditional distiller based on exergy output, exergy efficiency, and energy efficiency. The distillate yield, exergy efficiency, and energy efficiency of the modified SD were enhanced by 34 %, 49 %, and 40 %, respectively, compared with that of the traditional distiller. Moreover, the total amount of produced drinkable water using the modified SD could reach a high value of 6.1 L/day with a low cost of 0.014 $/L. © 2025

关键词： Exergy

来源：评论

学校读者我要写书评

暂无评论

Support Vector Machine based Multi-Class Classification for Oriented Instance Selection

Support Vector Machine based Multi-Class Classification for ...

引用

International Conference on Inventive Computation Technologies (ICICT)

作者： Sandeep Kumar Davuluri Deepak Srivastava Manisha Aeri Madhur Arora Ismail Keshta Richard Rivera Department of Information Technology University of the Cumberlands Williamsburg Kentucky Department of Computer Science and Engineering Himalayan school of Science and Technology Swami Rama Himalayan University Dehradun India CSE Graphic Era Hill University Dehradun India CSE Graphic Era deemed to be University Dehradun India RGPV University Bhopal Computer Science and Information Systems Department College of Applied Sciences AlMaarefa University Riyadh Saudi Arabia Department of Informatics and Computer Science Escuela Politécnica Nacional Quito Ecuador

Support vector machine has emerged as one of the most crucial classification techniques in recent years in pattern recognition, machine learning, and data mining. The model training will become more challenging when dealing with multi-classification problems, and its training time will increase considerably as the number of samples increases. A quick training data reduction technique MOIS that works with multiple classification problems is provided to address the issues above. This method takes the cluster centre as the reference point. While removing redundant training samples, selects the decisive boundary samples to significantly reduce the training data and reduce the distribution imbalance between categories. Experimental results show that MOIS can maintain or even improve the classification effect of support vector machines. Moreover, it can significantly improve training efficiency. For example, on the Opt digit data set, the proposed method improves the classification accuracy from 98.94% to 99.05% and, at the same time, shortens the training time to 15%; On the data set composed of the first 100 categories, when the accuracy rate is slightly improved (from 99.29% to 99.30%), the training time is significantly shortened to less than 6% of the original. In addition, MOIS itself has High operating efficiency.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Tutorial on Using Machine Learning and Deep Learning Models for Mental Illness Detection

arXiv

引用

arXiv 2025年

作者： Zhang, Yeyubei Wang, Zhongyan Ding, Zhanyi Tian, Yexin Dai, Jianglai Shen, Xiaorui Liu, Yunchong Cao, Yuchen University of Pennsylvania School of Engineering and Applied Science United States New York University Center for Data Science United States Georgia Institute of Technology College of Computing United States University of California Department of EECS Berkeley United States Northeastern University Khoury College of Computer Science United States

Social media has become an important source for understanding mental health, providing researchers a way to detect conditions like depression from user-generated posts. This tutorial provides practical guidance to address common challenges in applying machine learning and deep learning methods for mental health detection on these platforms. It focuses on strategies for working with diverse datasets, improving text preprocessing, and addressing issues such as imbalanced data and model evaluation. Real-world examples and step-by-step instructions demonstrate how to apply these techniques effectively, with an emphasis on transparency, reproducibility, and ethical considerations. By sharing these approaches, this tutorial aims to help researchers build more reliable and widely applicable models for mental health research, contributing to better tools for early detection and intervention. © 2025, CC BY.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Risk Assessment of Pregnancy-induced Hypertension Using a Machine Learning Approach 6

Risk Assessment of Pregnancy-induced Hypertension Using a Ma...

引用

Joint 6th International Conference on Digital Arts, Media and Technology with 4th ECTI Northern Section Conference on Electrical, Electronics, computer and Telecommunication engineering, ECTI DAMT and NCON 2021

作者： Wanriko, Sirinat Hnoohom, Narit Wongpatikaseree, Konlakom Jitpattanakul, Anuchit Musigavong, Olarik Mahidol University Image Information and Intelligence Laboratory Faculty of Engineering Department of Computer Engineering Nakhon Pathom Thailand Mahidol University Faculty of Engineering Department of Computer Engineering Nakhon Pathom Thailand King Mongkut's University of Technology North Bangkok Intelligent and Nonlinear Dynamic Innovations Research Center Faculty of Applied Science Department of Mathematics Bangkok Thailand Prachin Buri Thailand

ISBN: (纸本)9781665415699

This research aimed to develop a predictive model of the risk assessment of pregnancy-induced hypertension using a machine learning approach. Pregnancy-induced hypertension is a complication that has a serious impact on pregnant women and fetuses. It is the world's top three cause of death among pregnant women [1]. Nowadays, the exact cause of pregnancyinduced hypertension is unknown and therefore cannot be prevented. Early detection and received treatment can reduce the severity and danger. A public dataset of Logan (2020) was used in this research [2] the dataset was collected from a casecontrol study on the determinants of 83 pre-eclampsia and five eclampsia cases among 352 pregnant women delivering in county hospitals in Nairobi, Kenya. According to the dataset, 75 percent of the pregnant women were healthy. Only 25 percent of the pregnant women were pre-eclampsia and eclampsia. Thus, this would result in a problem of an imbalanced classification when one of the two classes had more data than the other class. As such, this problem was resolved with the Synthetic Minority Over-sampling Technique (SMOTE). Risk assessment of pregnancy-induced hypertension was performed on seven machine learning algorithms, which were logistic regression (LR), K-nearest neighbor (KNN), decision tree (DT), random forest (RF), multilayer perceptron neural network(MLP), support vector machines (SVM), and naive Bayes (NB). In the experimental results, RF had the highest accuracy at 89.62 percent compared to other machine learning algorithms. © 2021 IEEE.

关键词： Risk assessment

来源：评论

学校读者我要写书评

暂无评论

Creating and visualizing the materials science knowledge graph with whyis ?

Creating and visualizing the materials science knowledge gra...

引用

2021 International Semantic Web Conference Posters, Demos and Industry Tracks: From Novel Ideas to Industrial Practice, ISWC-Posters-Demos-Industry 2021

作者： McCusker, Jamie Deagen, Michael Fateye, Tolulomo Wallace, Anya Rashid, Sabbir M. McGuinness, Deborah L. Department of Computer Science Rensselaer Polytechnic Institute TroyNY United States College of Engineering and Mathematical Sciences University of Vermont BurlingtonVT United States Department of Mechanical Engineering and Materials Science Duke University DurhamNC United States Department of Mechanical Engineering and Applied Physics California Institute of Technology PasadenaCA United States

The NanoMine knowledge graph is being expanded to sup-port a broader range of materials, now including metamaterials, and is now called MaterialsMine. In the process, we have added new knowledge curation and visualization capabilities to Whyis, the framework that Ma-TerialsMine (and NanoMine before it) is built on. This demonstration will show how we use Whyis to support user-provided data uploads that con-form to the Dataset Catalog standard, use DOI and ORCiD to populate dataset metadata, curate data files into knowledge graph fragments us-ing Semantic Data Dictionaries, and allow domain scientists to visualize and explore the graph using SPARQL, Vega-Lite, and Data Voyager. © 2021 CEUR-WS. All rights reserved.

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

Comparison of Google Analytics with Similar Web for Statistical Analysis of Website Traffic 1st

Comparison of Google Analytics with Similar Web for Statisti...

引用

1st International Conference on Advanced Network Technologies and Computational Intelligence, ICANTCI 2024

作者： Garg, Bindu Kasar, Manisha Kotecha, Ketan Rahmani, Mohammad Khalid Imam University College of Engineering Pune India Symbiosis Center for Applied Artificial Intelligence Symbiosis Institute of Technology Symbiosis International University Pune India Department of Computer Science College of Computing and Informatics Saudi Electronic University Riyadh Saudi Arabia

ISBN: (纸本)9783031860713

This study addresses the burgeoning demand for website data collection and analysis in business operations, emphasizing the pivotal role of web analytics in providing crucial insights into customer behaviour. Despite the prevalence of various website traffic analysing tools, there exists a notable lack of comprehensive testing on the precision and reliability of these services. The research compares four key analytics measures from Google Analytics (GA) with SimilarWeb (SW) using average monthly statistics over a one-year period for 50+ websites across different nations and business verticals. Employing various statistical tests, including paired t-tests, Shapiro-Wilk tests, and correlation analysis, the study reveals statistically significant variations in key metrics between GA and SW. Notably, SimilarWeb tends to provide lower estimates for the number of visits and unique visitors compared to Google Analytics. However, the findings underscore the potential complementarity of using both tools in cases where direct access to a site’s analytics is not feasible, particularly for competitive intelligence and benchmarking purposes. The study’s extensive statistical analysis covers unique visitors, total visits, bounce rates, and average session duration, offering valuable insights into the nuanced variations and potential sources of error. There are statistically significant differences between the two web traffic tools in terms of unique visitors, total visits, average session time & bounce rates. In comparison to Google Analytics, SimilarWeb figures were on average, 40% less for unique visitors, 50% greater for average session time, 25% greater for bounce rate, and 20% lesser for total visits. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Error statistics

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：