检索结果-内蒙古大学图书馆

arXiv 2023年

作者： van Amsterdam, Wouter A.C. van Geloven, Nan Krijthe, Jesse H. Ranganath, Rajesh Cinà, Giovanni Department of Data Science and Biostatistics Julius Center of Health Sciences Primary Care University Medical Center Utrecht Heidelberglaan 100 Utrecht3584 CX Netherlands University of Utrecht Heidelberglaan 100 Utrecht3584 CX Netherlands Department of Biomedical Data Sciences Leiden University Medical Center Leiden Netherlands Pattern Recognition & Bioinformatics Delft University of Technology Delft Netherlands Courant Institute of Mathematical Science Department of Computer Science Center for Data Science New York University New York City United States Department of Medical Informatics Amsterdam University Medical Center Amsterdam Netherlands Institute for Logic Language and Computation University of Amsterdam Amsterdam Netherlands Pacmed Amsterdam Netherlands

Prediction models are popular in medical research and practice. By predicting an outcome of interest for specific patients, these models may help inform difficult treatment decisions, and are often hailed as the poster children for personalized, data-driven healthcare. We show however, that using prediction models for decision making can lead to harmful decisions, even when the predictions exhibit good discrimination after deployment. These models are harmful self-fulfilling prophecies: their deployment harms a group of patients but the worse outcome of these patients does not invalidate the predictive power of the model. Our main result is a formal characterization of a set of such prediction models. Next we show that models that are well calibrated before and after deployment are useless for decision making as they made no change in the data distribution. These results point to the need to revise standard practices for validation, deployment and evaluation of prediction models that are used in medical decisions. © 2023, CC BY.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

CIGAN: A Python Package for Handling Class Imbalance using Generative Adversarial Networks

arXiv

引用

arXiv 2022年

作者： Huang, Yuxiao Ma, Yan Data Science Columbian College of Arts & Sciences George Washington University WashingtonDC United States Department of Biostatistics and Bioinformatics Milken Institute School of Public Health George Washington University WashingtonDC United States

A key challenge in Machine Learning is class imbalance, where the sample size of some classes (majority classes) are much higher than that of the other classes (minority classes). If we were to train a classifier directly on imbalanced data, it is more likely for the classifier to predict a new sample as one of the majority classes. In the extreme case, the classifier could completely ignore the minority classes. This could have serious sociological implications in healthcare, as the minority classes are usually the disease classes (e.g., death or positive clinical test result). In this paper, we introduce a software that uses Generative Adversarial Networks to oversample the minority classes so as to improve downstream classification. To the best of our knowledge, this is the first tool that allows multi-class classification (where the target can have an arbitrary number of classes). The code of the tool is publicly available in our github repository (https://***/yuxiaohuang/research/tree/master/gwu/working/cigan/code). © 2022, CC BY.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Skew Probabilistic Neural Networks for Learning from Imbalanced data

SSRN

引用

SSRN 2024年

作者： Naik, Shraddha M. Chakraborty, Tanujit Hadid, Abdenour Chakraborty, Bibhas Manipal Academy of Higher Education Karnataka Manipal India Sorbonne Center for Artificial Intelligence Sorbonne University Abu Dhabi United Arab Emirates Centre for Quantitative Medicine Duke-NUS Medical School National University of Singapore Singapore Department of Statistics and Data Science National University of Singapore Singapore Department of Biostatistics and Bioinformatics Duke University United States

Real-world datasets often exhibit imbalanced data distribution, where certain class levels are severely underrepresented. In such cases, traditional pattern classifiers have shown a bias towards the majority class, impeding accurate predictions for the minority class. This paper introduces an imbalanced data-oriented approach using probabilistic neural networks (PNNs) with a skew normal probability kernel to address this major challenge. PNNs are known for providing probabilistic outputs, enabling quantification of prediction confidence and uncertainty handling. By leveraging the skew normal distribution, which offers increased flexibility, particularly for imbalanced and non-symmetric data, our proposed Skew Probabilistic Neural Networks (SkewPNNs) can better represent underlying class densities. To optimize the performance of the proposed approach on imbalanced datasets, hyperparameter fine-tuning is imperative. To this end, we employ a population-based heuristic algorithm, Bat optimization algorithms, for effectively exploring the hyperparameter space. We also prove the statistical consistency of the density estimates which suggests that the true distribution will be approached smoothly as the sample size increases. Experimental simulations have been conducted on different synthetic datasets, comparing various benchmark-imbalanced learners. Our real-data analysis shows that SkewPNNs substantially outperformstate-of-the-art machine learning methods for both balanced and imbalanced datasets in most experimental settings. © 2024, The Authors. All rights reserved.

关键词： Heuristic algorithms

来源：评论

学校读者我要写书评

暂无评论

LSD 4.0:an improved database for comparative studies of leaf senescence

引用

Molecular Horticulture 2022年第1期2卷 303-306页

作者： Jie Cao Yang Zhang Shuya Tan Qi Yang Hou‑Ling Wang Xinli Xia Jingchu Luo Hongwei Guo Zhang Zhang Zhonghai Li National Engineering Research Center for Tree Breeding and Ecological Restoration College of Biological Sciences and TechnologyBeijing Forestry UniversityBeijing 100083China National Genomics Data Center Beijing 100101China China National Center for Bioinformation Chinese Academy of SciencesBeijing 100101China University of Chinese Academy of Sciences Beijing 100049China College of Life Sciences Peking UniversityBeijing 100871China Center for Bioinformatics Peking UniversityBeijing 100871China Department of Biology Institute of Plant and Food ScienceSouthern University of Science and Technology(SUSTech)Shenzhen 518055GuangdongChina

Leaf senescence is the final stage of leaf development and involves the active degradation and dynamic transfer of its cellular components to newly growing and storage tissues,which contributes to plant fitness(Gan and Amasino,1997;Lim et al.,2007).The genetic modification of leaf senescence has emerged as a promising strategy for improving nutritional traits and stress tolerance in plants(Rivero et al.,2007).Efforts to dissect the molecular mechanisms underpinning leaf senescence reveal that it is a highly coordinated process regulated by a large number of senescence-associated genes(SAGs)(Guo and Gan,2005;Lim et al.,2007).Functional studies of these SAG genes through reverse genetics strategies and identification of senescence-altered mutants through forward genetic screening have deepened the understanding of leaf senescence(Guo et al.,2021).

关键词： traits senescence emerged

来源：评论

学校读者我要写书评

暂无评论

A Bayesian Network to Predict Cardiovascular Diseases in Adolescents and Young Adults with Breast Cancer

SSRN

引用

SSRN 2024年

作者： Bernasconi, Alice Zanga, Alessio Lucas, Peter J.F. Scutari, Marco Cosimo, Serena Di De Santis, Maria Carmen Rocca, Eliana La Baili, Paolo Cavallo, Ilaria Verderio, Paolo Ciniselli, Chiara M. Pizzamiglio, Sara Blanda, Adriana Perego, Paola Vallerio, Paola Stella, Fabio Trama, Annalisa Evaluative Epidemiology Unit Department of Epidemiology and Data Science Fondazione IRCCS Istituto Nazionale dei Tumori Milano Italy Department of Informatics Systems and Communication University of Milano-Bicocca Milano Italy Data Science and Advanced Analytics F. Hoffmann - La Roche Ltd Basel Switzerland University of Twente Enschede Netherlands Lugano Switzerland Department of Advanced Diagnostics Fondazione IRCCS Istituto Nazionale dei Tumori Milano Italy Radiation Oncology 1 Unit Fondazione IRCCS Istituto Nazionale dei Tumori Milano Italy Department of Radiation Oncology Azienda Ospedaliero Universitaria Integrata Verona Italy Data Science Unit Department of Epidemiology and Data Science Fondazione IRCCS Istituto Nazionale dei Tumori Milano Italy Scientific Direction Fondazione IRCCS Istituto Nazionale dei Tumori Milano Italy Unit of Bioinformatics and Biostatistics Department of Epidemiology and Data Science Fondazione IRCCS Istituto Nazionale dei Tumori Milano Italy Molecular Pharmacology Unit Department of Experimental Oncology Fondazione IRCCS Istituto Nazionale dei Tumori Milano Italy Cardiology Unit Fondazione IRCCS Istituto Nazionale dei Tumori Milano Italy

Background: When dealing with real-world healthcare data, data and knowledge fusion is crucial for providing accurate predictions. In this context, artificial intelligence (AI) tools such as Bayesian Networks (BNs) can be a perfect basis on which to build effective, interpretable models. Recently, the increasing number of adolescent and young adult (AYA) survivors of breast cancer (BC) has highlighted the cardiotoxic impact of cancer therapies in this peculiar and understudied age-group, making cardiovascular diseases (CVDs) among the most frequent long-term sequalae in these young patients. Against this background, we aimed to develop a predictive BN-based model to identify young BC survivors at higher risk of developing CVDs within 5-years from cancer ***: We trained and validated a BN model making use of expert knowledge and data a population-based (1036 patients) and a clinical (339 patient) cohort of female AYA (i.e., aged 18 to 39 years) 1-year survivors of BC, diagnosed in 2009-2014. We validated model performance against standard classification metrics and illustrated a possible model application for ranking patients by their risk probability using a lift ***: The proposed model showed very good classification performance (with all metrics above 69%) and a causal interpretable structure. Additionally, according to the lift chart, focusing on the 25% of 1-year BC survivors at higher risk of developing CVDs based on model predictions, 81% of those who would actually develop a CVD could be ***: In this paper, we developed and validated the first explainable model that can predict CVD risk in AYA survivors of BC. The model’s predictions can be used by clinicians to identify high-risk patients and efficiently rank patients by their predicted risk, to plan personalized follow-up ***: The research leading to these results has received funding from AIRC under IG 2020 - ID. 24864 – P.I. Annalisa ***

关键词： Bayesian networks

来源：评论

学校读者我要写书评

暂无评论

Position: Bayesian Deep Learning is Needed in the Age of Large-Scale AI 41

Position: Bayesian Deep Learning is Needed in the Age of Lar...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Papamarkou, Theodore Skoularidou, Maria Palla, Konstantina Aitchison, Laurence Arbel, Julyan Dunson, David Filippone, Maurizio Fortuin, Vincent Hennig, Philipp Hernández-Lobato, José Miguel Hubin, Aliaksandr Immer, Alexander Karaletsos, Theofanis Khan, Mohammad Emtiyaz Kristiadi, Agustinus Li, Yingzhen Mandt, Stephan Nemeth, Christopher Osborne, Michael A. Rudner, Tim G.J. Rügamer, David Teh, Yee Whye Welling, Max Wilson, Andrew Gordon Zhang, Ruqi Department of Mathematics The University of Manchester Manchester United Kingdom Eric and Wendy Schmidt Center Broad Institute of MIT and Harvard Cambridge United States Spotify London United Kingdom Computational Neuroscience Unit University of Bristol Bristol United Kingdom Centre Inria de l'Université Grenoble Alpes Grenoble France Department of Statistical Science Duke University United States Statistics Program KAUST Saudi Arabia Helmholtz AI Munich Germany Department of Computer Science Technical University of Munich Munich Germany Munich Center for Machine Learning Munich Germany Tübingen AI Center University of Tübingen Tübingen Germany Department of Engineering University of Cambridge Cambridge United Kingdom Department of Mathematics University of Oslo Oslo Norway Bioinformatics and Applied Statistics Norwegian University of Life Sciences Ås Norway Department of Computer Science ETH Zurich Switzerland Chan Zuckerberg Initiative CA United States Center for Advanced Intelligence Project RIKEN Tokyo Japan Vector Institute Toronto Canada Department of Computing Imperial College London London United Kingdom Department of Computer Science UC Irvine Irvine United States Department of Mathematics and Statistics Lancaster University Lancaster United Kingdom Department of Engineering Science University of Oxford Oxford United Kingdom Center for Data Science New York University New York United States Department of Statistics LMU Munich Munich Germany DeepMind London United Kingdom Department of Statistics University of Oxford Oxford United Kingdom Informatics Institute University of Amsterdam Amsterdam Netherlands Courant Institute of Mathematical Sciences Center for Data Science Computer Science Department New York University New York United States Department of Computer Science Purdue University West Lafayette United States

In the current landscape of deep learning research, there is a predominant emphasis on achieving high predictive accuracy in supervised tasks involving large image and language datasets. However, a broader perspective reveals a multitude of overlooked metrics, tasks, and data types, such as uncertainty, active and continual learning, and scientific data, that demand attention. Bayesian deep learning (BDL) constitutes a promising avenue, offering advantages across these diverse settings. This paper posits that BDL can elevate the capabilities of deep learning. It revisits the strengths of BDL, acknowledges existing challenges, and highlights some exciting research avenues aimed at addressing these obstacles. Looking ahead, the discussion focuses on possible ways to combine large-scale foundation models with BDL to unlock their full potential. Copyright 2024 by the author(s)

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Less is More: Reducing Overfitting in Deep Learning for EEG Classification

Less is More: Reducing Overfitting in Deep Learning for EEG ...

引用

Computers in Cardiology (CinC)

作者： Songchi Zhou Shijia Geng Jun Li Deyun Zhang Ziqian Xie Chuandong Cheng Shenda Hong Department of Bioinformatics and Biostatistics Shanghai Jiao Tong University Shanghai China HeartVoice Medical Technology Hefei China College of Electronic Science and Engineering Jilin University Changchun China University of Texas Health Science Center Houston TX US Department of Neurosurgery The First Affiliated Hospital of USTC Hefei China National Institute of Health Data Science Peking University Beijing China Institute of Medical Technology Health Science Center of Peking University Beijing China

Although most of the patients' recordings includes large scale long-term physiological time series, the patient-level quantity is relatively small, posing great challenges for machine learning methods. As part of the George B. Moody PhysioNet Challenge 2023, we aim to propose a series of Reducing Overfitting techniques in Deep learning for EEG (coined as RODE) in this scenario, for neurological recovery prognosis. RODE is a simple yet effective machine-learning method, which is mostly powered by generalizable deep-learning features. Specifically, we first pre-train a convolutional neural network on the segment level with a margin-based and mining-based loss, and extract deep features from it. Then we decrease the deep features' complexity using dimensionality reduction methods, which prove to be quite significant for reducing overfitting in deep learning. Finally, we combine the reduced features with static features in patient level, and put them into an ensemble model for classification. Our team, PKU_NIHDS, receives a Challenge score of 0.821 on the hidden validation set and 0.708 on the hidden test set.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multivariate Time-Series Deep Learning for Joint Prediction of Temperature and Relative Humidity in a Closed Space

引用

Procedia Computer science 2023年 227卷 1046-1053页

作者： Fergianto E. Gunawan Arief S. Budiman Bens Pardamean Endang Juana Sugiarto Romeli Tjeng W. Cenggoro Kartika Purwandari Alam A. Hidayat Anak. A.N.P. Redi Muhammad Asrol Industrial Engineering Department BINUS Graduate Program – Master of Industrial Engineering Bina Nusantara University Jakarta 11480 Indonesia Computer Science Department BINUS Graduate Program – Master of Computer Science Program Bina Nusantara University Jakarta 11480 Indonesia Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta 11480 Indonesia Electrical Engineering Department Faculty of Industrial Technology Universitas Trisakti Jakarta 11440 Indonesia PT Impack Pratama Industri Jakarta 14350 Indonesia Computer Science Department School of Computer Science Bina Nusantara University Jakarta 11480 Indonesia Mathematics Department School of Computer Science Bina Nusantara University Jakarta 11480 Indonesia

An accurate predictive model of temperature and humidity plays a vital role in many industrial processes that utilize a closed space such as in agriculture and building management. With the exceptional performance of deep learning on time-series data, developing a predictive temperature and humidity model with deep learning is propitious. In this study, we demonstrated that deep learning models with multivariate time-series data produce remarkable performance for temperature and relative humidity prediction in a closed space. In detail, all deep learning models that we developed in this study achieve almost perfect performance with an R value over 0.99.

关键词： closed space deep learning humity prediction indoor temperature prediction

来源：评论

学校读者我要写书评

暂无评论

database management system design for automatic weather information with twitter data collection 6

Database management system design for automatic weather info...

引用

6th International Conference on Information Management and Technology, ICIMTech 2021

作者： Purwandari, Kartika Perbangsa, Anzaludin Samsinga Sigalingging, Join W. C. Krisna, Adhi Asta Anggrayani, Suci Pardamean, Bens Bina Nusantara University Bioinformatics and Data Science Research Center Jakarta11480 Indonesia Bina Nusantara University School of Information Systems Information Systems Department Jakarta11480 Indonesia Meteorological Climatological and Geophysical Agency Database Center Division of Bmkg Jakarta10720 Indonesia Bina Nusantara University Binus Graduate Program - Master of Computer Science Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9781665449373

Improvement of data management about weather in Indonesia is very supportive of the need for weather information. Currently, the weather information that is disseminated to the public is only based on sensor location points that are spread across Indonesia. However, supporting data generated from these sensors is only limited to certain areas. We hence develop a database system that can help collect weather information automatically via Twitter. This facilitates the Indonesian people in knowing weather information based on the completeness of data from sensor data and Twitter. The steps from the database that are built will be very helpful in collecting data about the weather throughout Indonesia. The basic step that we use in our database are conceptual, logical and physical design. The information obtained in this study is very important for efficient and effective data collection to achieve real-time automation of weather information. © 2021 IEEE.

关键词： database systems

来源：评论

学校读者我要写书评

暂无评论

Network security modeling using NetFlow data: Detecting botnet attacks in IP traffic

arXiv

引用

arXiv 2021年

作者： Subramaniam, Ganesh Chen, Huan Varadhan, Ravi Archibald, Robert Data Science and AI Research At and T Chief Data Office Department of Biostatistics Johns Hopkins Bloomberg School of Public Health Division of Biostatistics and Bioinformatics Department of Oncology Johns Hopkins University

Cybersecurity, security monitoring of malicious events in IP traffic, is an important field largely unexplored by statisticians. Computer scientists have made significant contributions in this area using statistical anomaly detection and other supervised learning methods to detect specific malicious events. In this research, we investigate the detection of botnet command and control (C&C) hosts in massive IP traffic. We use the NetFlow data, the industry standard for monitoring of IP traffic for exploratory analysis and extracting new features. Using statistical as well as deep learning models, we develop a statistical intrusion detection system (SIDS) to predict traffic traces identified with malicious attacks. Employing interpretative machine learning techniques, botnet traffic signatures are derived. These models successfully detected botnet C&C hosts and compromised devices. The results were validated by matching predictions to existing blacklists of published malicious IP addresses. Copyright © 2021, The Authors. All rights reserved.

关键词： Network security

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：