检索结果-内蒙古大学图书馆

A pretrained transformer model for decoding individual glucose dynamics from continuous glucose monitoring data

National science Review 2025年第05期 362-381页

作者： Yurun Lu Dan Liu Zhongming Liang Rui Liu Pei Chen Yitong Liu Jiachen Li Zhanying Feng Lei M.Li Bin Sheng Weiping Jia Luonan Chen Huating Li Yong Wang Center for Excellence in Mathematical Sciences National Center for Mathematics and Interdisciplinary Sciences Hua Loo-Keng Center for Mathematical Sciences Key Laboratory of Management Decision and Information System Academy of Mathematics and Systems ScienceChinese Academy of Sciences School of Mathematics University of Chinese Academy of SciencesChinese Academy of Sciences Department of Endocrinology and Metabolism Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of MedicineShanghai Diabetes InstituteShanghai Clinical Center for Diabetes Shanghai Key Laboratory of Diabetes Mellitus Key Laboratory of Systems Health Science of Zhejiang Province School of Life ScienceHangzhou Institute for Advanced StudyUniversity of Chinese Academy of Sciences BGI-Research School of Mathematics South China University of Technology Department of Statistics Department of Biomedical Data Science Bio-X ProgramStanford University Department of Computer Science and Engineering Shanghai Jiao Tong University State Key Laboratory of Cell Biology Center for Excellence in Molecular Cell Science Shanghai Institute of Biochemistry and Cell Biology Chinese Academy of Sciences Guangdong Institute of Intelligence Science and Technology Pazhou Laboratory (Huangpu)

Continuous glucose monitoring(CGM) technology has grown rapidly to track real-time blood glucose levels and trends with improved sensor accuracy. The ease of use and wide availability of CGM will facilitate safe and effective decision making for diabetes management. Here, we developed an attention-based deep learning model, CGMformer, pretrained on a well-controlled and diverse corpus of CGM data to represent individual's intrinsic metabolic state and enable clinical applications. During pretraining, CGMformer encodes glucose dynamics including glucose level, fluctuation, hyperglycemia, and hypoglycemia into latent space with self-supervised learning. It shows generalizability in imputing glucose value across five external datasets with different populations and metabolic states(MAE = 3.7 mg/d L). We then fine-tuned CGMformer towards a diverse panel of downstream tasks in the screening of diabetes and its complications using task-specific data, which demonstrated a consistently boosted predictive accuracy over direct fine-tuning on a single task(AUROC = 0.914 for type 2 diabetes(T2D) screening and 0.741 for complication screening). By learning an intrinsic representation of an individual's glucose dynamics,CGMformer classifies non-diabetic individuals into six clusters with elevated T2D risks, and identifies a specific cluster with lean body-shape but high risk of glucose metabolism disorders, which is overlooked by traditional glucose measurements. Furthermore, CGMformer achieves high accuracy in predicting an individual's postprandial glucose response with dietary modelling(Pearson correlation coefficient = 0.763)and helps personalized dietary recommendations. Overall, CGMformer pretrains a transformer neural network architecture to learn an intrinsic representation by borrowing information from a large amount of daily glucose profiles, and demonstrates predictive capabilities fine-tuned towards a broad range of downstream applications, holding promise for the ear

关键词：

来源：评论

学校读者我要写书评

暂无评论

Increasing accuracy of classification physical activity based on smartphone using ensemble logistic regression with boosting method 3

Increasing accuracy of classification physical activity base...

引用

3rd International Conference on science, ICOS 2019

作者： Lawi, A. Aziz, F. Wungo, S.L. Department of Computer Science Universitas Hasanuddin Makassar90245 Indonesia Faculty of Mathematics and Natural Sciences Universitas Pancasakti Makassar90132 Indonesia Program Study Informatics Engineering STMIK Kreatindo Monokwari98300 Indonesia

At present, the smartphone is equipped with several sensors such as Accelerometer, Gravity sensor, and Gyroscope which can be used to recognize human physical activities such as walking upstair and walking downstairs etc. Machine learning is needed to group data and get information. Statistical methods have poor performance in classifying because procedures must be met. To overcome this, an ensemble technique was used. This study proposes the application of the gradientboost ensemble method to classify walking upstair and walking downstairs. The *** system is designed for data retrieval using a smartphone. Then, the dataset will be partitioned into 70% training data and 30% test data. The results show that the performance of the ensemble boosting method produces 81.82% accuracy, 86.11% sensitivity and 77.50% specificity. © Published under licence by IOP Publishing Ltd.

关键词： Smartphones

来源：评论

学校读者我要写书评

暂无评论

Application of XGBoost algorithm in classification of 28 types of cancer based on ctDNA mutations

引用

AIP Conference Proceedings 2024年第1期2970卷

作者： Favorisen R. Lumbanraja Rachelia Dita Anggraini Yohana Tri Utami Rosma Hasibuan Computer Science Study Program Faculty of Mathematics and Natural Sciences University of Lampung Jl. Prof. Dr. Sumantri Brojonegoro No. 1 Bandar Lampung 35145 Indonesia Department of Agrotechnology Faculty of Agriculture University of Lampung Jl. Prof. Dr. Sumantri Brojonegoro No. 1 Bandar Lampung 35145 Indonesia

Circulating Tumor DNA (ctDNA) in blood samples of cancer patients contains components for the diagnosis of cancer. These components when combined with bioinformatics using machine learning algorithms can facilitate the classification of cancer. Extreme gradient boosting (XGBoost) is an ensemble learning method that works by combining basic models to produce an optimal predictive model. This study aims to determine the results of the XGBoost algorithm classification using somatic mutations and copy number changes that occur in ctDNA. This study performs a classification using XGBoost with two types of cross-validation, 10-fold and holdout cross-validation, using data from Soh et al., 2017 [1]. For holdout cross-validation, training and testing data will be divided into 60:40, 70:30, 75:25, 80:20, and 90:10. Each classification experiment conducted in this study was tested using two types of data, namely data with overall features and data with the top 900 features as a result of feature selection. The highest accuracy result was obtained by XGBoost holdout cross-validation 90:10 using 900 features with an accuracy value of 89.91%, while the lowest accuracy value was obtained by XGBoost 10-fold cross-validation 900 features, which was 70.01%. The accuracy value tends to be higher in the 900-feature experiment compared to the overall feature. The holdout cross-validation experiment gets a better accuracy value than the 10-fold cross-validation experiment. For holdout cross-validation, the greater the percentage of training data, the higher the accuracy obtained.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Galaxy clusters at z ∼ 1 imaged by ALMA with the Sunyaev–Zel’dovich effect

引用

Publications of the Astronomical Society of Japan 2023年第2期75卷 311-337页

作者： Kitayama, Tetsu Ueda, Shutaro Okabe, Nobuhiro Akahori, Takuya Hilton, Matt Hughes, John P. Ichinohe, Yuto Kohno, Kotaro Komatsu, Eiichiro Lin, Yen-Ting Miyatake, Hironao Oguri, Masamune Sifón, Cristóbal Takakuwa, Shigehisa Takizawa, Motokazu Tsutsumi, Takahiro van Marrewijk, Joshiwa Wollack, Edward J. Department of Physics Toho University 2-2-1 Miyama Chiba Funabashi274-8510 Japan No. 1 Section 4 Roosevelt Road Taipei10617 Taiwan Institute of Astronomy National Tsing Hua University Hsinchu30013 Taiwan Department of Physics Hiroshima University 1-3-1 Kagamiyama Higashi-Hiroshima Hiroshima739-8526 Japan Astrophysical Science Center Hiroshima University 1-3-1 Kagamiyama Higashi-Hiroshima Hiroshima739-8526 Japan Core Research for Energetic Universe Department of Physics Hiroshima University 1-3-1 Kagamiyama Higashi-Hiroshima Hiroshima739-8526 Japan Physics Program Graduate School of Advanced Science and Engineering Hiroshima University 1-3-1 Kagamiyama Higashi-Hiroshima Hiroshima739-8526 Japan Mizusawa VLBI observatory National Astronomical Observatory of Japan 2-21-1 Osawa Mitaka Tokyo181-8588 Japan Operation Division SKA Observatory Jodrell Bank Lower Withington MacclesfieldSK11 9DL United Kingdom Astrophysics Research Centre University of KwaZulu-Natal Westville Campus Durban4041 South Africa School of Mathematics Statistics & Computer Science University of KwaZulu-Natal Westville Campus Durban4041 South Africa Wits Centre for Astrophysics School of Physics University of the Witwatersrand Private Bag 3 Johannesburg2050 South Africa Department of Physics and Astronomy Rutgers The State University of New Jersey PiscatawayNJ08854-8019 United States Department of Physics Rikkyo University 3-34-1 Nishi-Ikebukuro Toshima-ku Tokyo171-8501 Japan Institute of Astronomy School of Science The University of Tokyo 2-21-1 Osawa Mitaka Tokyo181-0015 Japan Research Center for the Early Universe School of Science The University of Tokyo 7-3-1 Hongo Bunkyo Tokyo113-0033 Japan Max-Planck-Institut für Astrophysik Karl-Schwarzschild Str. 1 GarchingD-85741 Germany The University of Tokyo Institutes for Advanced Study The University of Tokyo 5-1-5 Kashiwanoha Chiba Kashiwa277-8583 Japan Nagoya University Furo-cho Chikusa-

We present high angular resolution measurements of the thermal Sunyaev–Zel’dovich effect (SZE) toward two galaxy clusters, RCS J2319+0038 at z = 0.9 and HSC J0947−0119 at z = 1.1, by the Atacama Large Millimeter/submillimeter Array (ALMA) in Band 3. They are supplemented with available Chandra X-ray data, optical data taken by Hyper Suprime-Cam on Subaru, and millimeter-wave SZE data from the Atacama Cosmology Telescope. Taking into account departures from spherical symmetry, we have reconstructed non-parametrically the inner pressure profile of two clusters as well as electron temperature and density profiles for RCS J2319+0038. This is one of the first such measurements for an individual cluster at z 0.9. We find that the inner pressure profile of both clusters is much shallower than that of local cool-core clusters. Our results consistently suggest that RCS J2319+0038 hosts a weak cool core, where radiative cooling is less significant than in local cool cores. On the other hand, HSC J0947−0119 exhibits an even shallower pressure profile than RCS J2319+0038 and is more likely to be a non-cool-core cluster. The SZE centroid position is offset by more than 140 h−170 kpc from the peaks of galaxy distribution in HSC J0947−0119, suggesting a stronger influence of mergers in this cluster. We conclude that these distant clusters are at a very early stage of developing the cool cores typically found in clusters at lower redshifts. © The Author(s) 2023. Published by Oxford University Press on behalf of the Astronomical Society of Japan.

关键词： Millimeter waves

来源：评论

学校读者我要写书评

暂无评论

Large language models for diabetes training:a prospective study

引用

science Bulletin 2025年第6期70卷 934-942页

作者： Haoxuan Li Zehua Jiang Zhouyu Guan Yuqian Bao Yuexing Liu Tingting Hu Jiajia Li Ruhan Liu Liang Wu Di Cheng Hongwei Ji Yong Wang Ya-Xing Wang Carol Y.Cheung Yingfeng Zheng Jihong Wang Zhen Li Weibing Wu Cynthia Ciwei Lim Yong Mong Bee Hong Chang Tan Elif I.Ekinci David C.Klonoff Justin B.Echouffo-Tcheugui Nestoras Mathioudakis Leonor Corsino Rafael Simó Charumathi Sabanayagam Gavin Siew Wei Tan Ching-Yu Cheng Tien Yin Wong Huating Li Chun Cai Lijuan Mao Lee-Ling Lim Yih-Chung Tham Bin Sheng Weiping Jia Shanghai University of Sport Shanghai 200438China School of Clinical Medicine Beijing Tsinghua Changgung HospitalTsinghua MedicineTsinghua UniversityBeijing 100084China Beijing Visual Science and Translational Eye Research Institute(BERI) Tsinghua MedicineTsinghua UniversityBeijing 100084China Shanghai Belt and Road International Joint Laboratory of Intelligent Prevention and Treatment for Metabolic Diseases Department of Computer Science and EngineeringSchool of ElectronicInformationand Electrical EngineeringShanghai Jiao Tong UniversityDepartment of Endocrinology and MetabolismShanghai Sixth People’s Hospital Affiliated to Shanghai Jiao Tong University School of MedicineShanghai Diabetes InstituteShanghai Clinical Center for DiabetesShanghai 200240China MoE Key Lab of Artificial Intelligence Artificial Intelligence InstituteShanghai Jiao Tong UniversityShanghai 200240China Department of Cardiology the Affiliated Hospital of Qingdao UniversityQingdao 266011China CEMS NCMISHCMSMADISAcademy of Mathematics and Systems ScienceChinese Academy of SciencesBeijing 100190China School of Mathematics University of Chinese Academy of SciencesChinese Academy of SciencesBeijing 100049China Key Laboratory of Systems Health Science of Zhejiang Province School of Life ScienceHangzhou Institute for Advanced StudyUniversity of Chinese Academy of SciencesHangzhou 310024China Beijing Institute of Ophthalmology Beijing Tongren HospitalCapital Medical UniversityBeijing Ophthalmology and Visual Sciences Key LaboratoryBeijing 100730China k.Department of Ophthalmology and Visual Sciences The Chinese University of Hong KongHong Kong 999077China State Key Laboratory of Ophthalmology Zhongshan Ophthalmic CenterSun Yat-sen UniversityGuangdong Provincial Key Laboratory of Ophthalmology and Visual ScienceGuangdong Provincial Clinical Research Center for Ocular DiseasesGuangzhou 510060China Department of Renal Medicine Singapore General HospitalSingHealth-Duke Academic Medical CentreSingapore169

Diabetes poses a considerable global health challenge,with varying levels of diabetes knowledge among healthcare professionals,highlighting the importance of diabetes *** Language Models(LLMs)provide new insights into diabetes training,but their performance in diabetes-related queries remains uncertain,especially outside the English language like *** first evaluated the performance of ten LLMs:ChatGPT-3.5,ChatGPT-4.0,Google Bard,LlaMA-7B,LlaMA2-7B,Baidu ERNIE Bot,Ali Tongyi Qianwen,MedGPT,HuatuoGPT,and Chinese LlaMA2-7B on diabetes-related queries,based on the Chinese National Certificate Examination for Primary Diabetes Care in China(NCE-CPDC)and the English Specialty Certificate Examination in Endocrinology and Diabetes of Membership of the Royal College of Physicians of the United ***,we assessed the training of primary care physicians(PCPs)without and with the assistance of ChatGPT-4.0 in the NCE-CPDC examination to ascertain the reliability of LLMs as medical *** found that ChatGPT-4.0 outperformed other LLMs in the English examination,achieving a passing accuracy of 62.50%,which was significantly higher than that of Google Bard,LlaMA-7B,and *** the NCE-CPFC examination,ChatGPT-4.0,Ali Tongyi Qianwen,Baidu ERNIE Bot,Google Bard,MedGPT,and ChatGPT-3.5 successfully passed,whereas LlaMA2-7B,HuatuoGPT,Chinese LLaMA2-7B,and LlaMA-7B ***-4.0(84.82%)surpassed all PCPs and assisted most PCPs in the NCE-CPDC examination(improving by 1%–6.13%).In summary,LLMs demonstrated outstanding competence for diabetes-related questions in both the Chinese and English language,and hold great potential to assist future diabetes training for physicians globally.

关键词： Diabetes Diabetestraining Largelanguagemodels Primarydiabetescare Prospectivestudy

来源：评论

学校读者我要写书评

暂无评论

An analysis of Fama and French three factor model in market reaction to Indonesia presidential election in 2019 3

An analysis of Fama and French three factor model in market ...

引用

3rd International Conference on science, ICOS 2019

作者： Diaraya Haryanto, L. Lawi, A. Amir, S. Navira, A. Raya, A.R. Computer Science Study Program Faculty of Mathematics and Natural Sciences Hasanuddin University Indonesia Department of Mathematics Faculty of Mathematics and Natural Sciences Hasanuddin University Indonesia Department of Accounting Faculty of Economy and Business Hasanuddin University Indonesia

This study reveals stock market reaction to Indonesia presidential election in 2019 using Fama and French Three Factor Model with accounting dependent variable being Average Abnormal Return (AAR) in addition to Earning Per Share (EPS), Debt to Equity ratio (DER), Return on Assets (ROA), and Return on Equity (ROE) as independent variables. The research population was all stocks of the companies registered in the Indonesia Stock Exchange in 2019 with their dividend announcement. The characteristic of the observed object was stock price measured by return, abnormal return and average abnormal return during 111-day-observation comprising of 11 days of windows period observation and 100 days of estimation period. The data from 101 companies as samples determined through the purposive sampling technique were analyzed using SPSS software version 23, which represented all industrial sectors in Indonesia Stock Exchange. The measure and analysis of companies stocks concluded that a gradual rise in AAR values was witnessed in all stock portfolios during the announcement date. © Published under licence by IOP Publishing Ltd.

关键词： Financial markets

来源：评论

学校读者我要写书评

暂无评论

Review the Utilization of Big Data and K-Means Algorithm in Supporting The Determination of Village Status As Support To The Ministry of Village PDTT

引用

Journal of Physics: Conference Series 2021年第1期1811卷

作者： Paska Marto Hasugian Harvei Desmon Hutahaean Bosker Sinaga Sriadhi Saronom Silaban Informatics Engineering Study Program STMIK Pelita Nusantara Medan Indonesia Computer Science and Informatics Education Study Program Faculty of Engineering Universitas Negeri Medan Medan 20221 Indonesia Departement of chemistry Universitas Negeri Medan Medan 20221 Indonesia

Big data refers to big data, fast data processing, diversity of data structures, and data values so that it is not possible to be processed with outdated methods. Big data technology is used in various industrial sectors. Big data technology is the whole technology that can handle Big Data. Some of the uses of big data are based on the largest significant data traffic sources such as social media, financial transactions, public data, sensor data, and corporate data. The same is the case with the status of so many villages in Indonesia, so it is better to use big data for the classification of village status based on the village index build by involving an algorithm process. This research aims to produce a description of the role of big data in supporting activities in a grouping. The method used is a qualitative approach related to data collection based on scientific work with the source of data information needed to study literature techniques from various studies that have been published in national and international journals. A decision is made that big data has been widely used in different circles to facilitate performance and speed up the decision-making process.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Design of big data technology prototype for classification of village status based on village development index involves k-means algorithm to support village ministry Pdtt work programs

引用

Journal of Physics: Conference Series 2021年第1期1811卷

Big data technology is the overall technology that can handle the processing associated with analyzing the data to explore the potential that is in it. Some of the uses of big data are based on the biggest data traffic sources, such as social media, financial transactions, public data, censorship data, and company data. The problem found in the accumulation of village status data that has not been utilized in the decision-making process to determine the status of the next village and the absence of an application to process the data stored in the data database so that a tool is developed to analyze the buildup. The results of this study are to produce a prototype design that is Unified Modeling Language (UML) design, Form Design, and Database Design that will be used to develop Big Data Technology in determining the classification of village status based on the Village Build index.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Publication of the society's sensitive personal data in the legitimate site of the general election committee

引用

Journal of Physics: Conference Series 2020年第1期1516卷

作者： A Parwata K Y E Aryanto D G H Divayana Study Program of Computer Science Postgraduate Program Ganesha University of Education Singaraja Indonesia

It is highly important to protect the society's sensitive personal data. The reason is that many personal data are published by the government institution without paying attention to the prevailing regulations. In this current study, the search for the data in the form of hyperlink with its contents in the legitimate site of the General Election Committees was conducted with the assistance of Crawling Web. The obtained contents were preprocessed using the Preprocessing Text, which were then weighted using the TF-IDF method before they were classified using the Naїve Bayes method. After that, an analysis on the types of the published sensitive personal and the extent of the publication based on the area groups was conducted. Out of 6,700 instances of the personal data which were analyzed, 6.430 were published. The personal data which were published were full name, place and date of birth, religion, marital status, ID Number of the government civil servants, identity card number, number of the tax payer, account number, mobile number, e-mail, address, position, and face photo. The level of publication based on the total data found was as follows: 11.45% in the Central General Election Committee, 21.60% in the eastern area, 17.01% in the central area and 49.94% in the eastern area. The accuracy of the Naїve Bayes method averaged 96.99%. Prior to publication, the General Election Committee is recommended to respect someone's personal data as privacy and the data which would be published should obtain approval and an easily-contacted contact person.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Optimization of artificial neural networks to improve accuracy of vocational competence selection of vocational school students using nguyen-widrow

引用

Journal of Physics: Conference Series 2020年第1期1516卷

作者： I G P Christyaditama I M Candiasa I G A Gunadi Study Program of Computer Science Postgraduate Program Ganesha University of Education Singaraja Indonesia

The present study used Neural Network Backpropagation combined with Nguyen-Widrow to optimize the disadvantages of ANN causes, which is the difficulty in initializing initial weights. The test was conducted on a dataset of values in semesters 1 and 2. The test results show that the best performance in training model of artificial neural networks with Nguyen-widrow is the smallest average MSE error of 0.002 and the highest average accuracy of 96.38%. Training on artificial neural network model training data with Nguyen-widrow has the smallest MSE error, 0.000996 and the highest accuracy is 97.49% on architecture ANN 9-9.1 with training function parameters: traingdx, epoch: 1000, learning rate: 0.1, and error: 0.001. The best performance was also seen in testing the testing of artificial neural network models with Nguyen-widrow with the smallest average error-MSE of 0.026 and the highest average accuracy of 87.85%. Training data testing on artificial neural network models with Nguyen-widrow has the smallest error-MSE which is 0.004436 and the highest accuracy is 94.50% on architecture ANN 9-9.1 with training function parameters: traingdx, epoch: 1000, learning rate: 0.1, and error: 0.001. The artificial neural network model with Nguyen-widrow has an accuracy difference of 8.53% smaller than the artificial neural network model with an accuracy difference of 9.28%. It can be concluded that the Artificial Neural Network with Nguyen-Widrow can overcome the ANN problem in determining initial weights so that it gives an increase in the accuracy of the prediction of students' competency selection better than the Artificial Neural Network without Nguyen-Widrow.

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：