检索结果-内蒙古大学图书馆

41st International Conference on machine learning, ICML 2024

作者： Dao, Tri Gu, Albert Department of Computer Science Princeton University United States Machine Learning Department Carnegie Mellon University United States

While Transformers have been the main architecture behind deep learning's success in language modeling, state-space models (SSMs) such as Mamba have recently been shown to match or outperform Transformers at small to medium scale. We show that these families of models are actually quite closely related, and develop a rich framework of theoretical connections between SSMs and variants of attention, connected through various decompositions of a well-studied class of structured semiseparable matrices. Our state space duality (SSD) framework allows us to design a new architecture (Mamba-2) whose core layer is an a refinement of Mamba's selective SSM that is 2-8× faster, while continuing to be competitive with Transformers on language modeling. Copyright 2024 by the author(s)

关键词： Distribution transformers

来源：评论

学校读者我要写书评

暂无评论

THE HUMAN-AI SUBSTITUTION GAME: ACTIVE learning FROM A STRATEGIC LABELER 12

THE HUMAN-AI SUBSTITUTION GAME: ACTIVE LEARNING FROM A STRAT...

引用

12th International Conference on learning Representations, ICLR 2024

作者： Yan, Tom Zhang, Chicheng Machine Learning Department Carnegie Mellon University United States Department of Computer Science University of Arizona United States

The standard active learning setting assumes a willing labeler, who provides labels on informative examples to speed up learning. However, if the labeler wishes to be compensated for as many labels as possible before learning finishes, the labeler may benefit from actually slowing down learning. This incentive arises for instance if the labeler is to be replaced by the ML model once it is trained. In this paper, we initiate the study of learning from a strategic labeler, who may abstain from labeling to slow down learning. We first prove that strategic abstention can prolong learning, and propose a novel complexity measure and representation to analyze the query complexity of the learning game. Next, we develop a near-optimal deterministic algorithm, prove its robustness to strategic labeling, and contrast it with other active learning algorithms. We also analyze extensions that encompass more general learning goals and labeler assumptions. Finally, we characterize the query cost of multi-task active learning, with and without abstention. Our first exploration of strategic labeling aims to consolidate our theoretical understanding of the imitative nature of ML in human-AI interaction. © 2024 12th International Conference on learning Representations, ICLR 2024. All rights reserved.

关键词： learning algorithms

来源：评论

学校读者我要写书评

暂无评论

UNCCER: Unified Network for Cancer Classification and Efficient Representation using Microarray data 18

UNCCER: Unified Network for Cancer Classification and Effici...

引用

18th International Conference on Open Source Systems and Technologies, ICOSST 2024

作者： Younis, Haseeb Brahmi, Horiya Imane Byrne, Jonathan Minghim, Rosane University College Cork Department of Computer Science Ireland Intel Ireland Limited Machine Learning Group Ireland

ISBN: (纸本)9798331508692

In recent years, there has been a significant advance in the use of machine learning (ML) techniques to extract gene expression data from microarray databases, particularly in cancer-related research. There no unified method for classifying cancer microarray data, even after ML adoption. Due to the high dimensionality of microarray data, it is difficult to extract the relevant features and provide insights that can be helpful in identifying cancer types and stages. In this paper, we propose a Unified Network for Cancer Classification and Efficient Representation (UNCCER) using Deep learning (DL) on cancer microarray data. To implement this methodology, we employed a microarray database (CuMiDa) that has 78 carefully curated datasets for different types of cancers. Our single model has the capability to learn the patterns, cluster instances into their corresponding classes, and classify the cancer. We also used the Uniform Manifold Approximation and Projection (UMAP) to visualise, in low dimension, the instance separation both on original data and transformed data by our methodology. Using the proposed methodology, we achieved average 94% average accuracy, precision, recall, F1 Score, and 91% G-Mean. © 2024 IEEE.

关键词： Lung cancer

来源：评论

学校读者我要写书评

暂无评论

Reducing sequential change detection to sequential estimation 41

Reducing sequential change detection to sequential estimatio...

引用

41st International Conference on machine learning, ICML 2024

作者： Shekhar, Shubhanshu Ramdas, Aaditya Department of Statistics and Data Science Carnegie Mellon University United States Machine Learning Department Carnegie Mellon University United States

We consider the problem of sequential change detection under minimal assumptions on the distribution generating the stream of observations. Formally, our goal is to design a scheme for detecting any changes in a parameter or functional θ of the data stream distribution that has small detection delay, but guarantees control on the frequency of false alarms in the absence of changes. We describe a simple reduction from sequential change detection to sequential estimation using confidence sequences (CSs): begin a new level-(1 − α) CS at each time step, and proclaim a change as soon as the intersection of all active CSs becomes empty. We prove that the average run length of our scheme is at least 1/α, resulting in a change detection scheme with minimal structural assumptions (thus allowing for possibly dependent observations, and nonparametric distribution classes), but strong guarantees. We also describe an interesting parallel with Lorden's reduction from change detection to sequential testing and connections to the recent "e-detector" framework. Copyright 2024 by the author(s)

关键词：

来源：评论

学校读者我要写书评

暂无评论

Design of Weed Detection Technique Using Yolo Models

Design of Weed Detection Technique Using Yolo Models

引用

2024 IEEE Students Conference on Engineering and Systems, SCES 2024

作者： Sreehari, M. Jain, Ankur Vellore Institiute of Technology Artificial intelligence and Machine Learning Department of computer science engineering specialization Bhopal India

ISBN: (纸本)9798350374711

The use of the ADAM (Adaptive Moment Estimation) and SGD (Stochastic Gradient Descent) algorithms to optimize the YOLOv7(You Only Look Once), YOLOv8, and YOLO-NAS models for weed detection in agricultural landscapes is investigated in this study. Offering a thorough comparative study based on mean Average Precision (mAP), recall, and precision measures is the goal. When compared to YOLOv7, YOLOv8 regularly performs better, exhibiting balanced recall and higher precision. With remarkable mAP scores at IOU (Intersection Over Union) 0.50, YOLOv8 ADAM in particular sticks out, demonstrating strong weed identification abilities. The YOLOv7 ADAM exhibits a precision-recall trade-off, with precision being slightly sacrificed for recall. Model superiority is not solely determined by the optimization algorithm used (ADAM vs. SGD);YOLOv8 variations perform admirably. For real-world implementation, practical factors like processing efficiency and deployment viability are even more important than quantitative criteria. This work establishes the foundation for comprehending the complex trade-offs and advantages that are present in every model and algorithm, in addition to helping with the educated selection of the best YOLO models for weed detection. Future research directions in object detection in agriculture, including ensemble techniques and hyperparameter tuning, promise more advancements in this subject as technology develops. © 2024 IEEE.

关键词： Stochastic systems

来源：评论

学校读者我要写书评

暂无评论

Dynamic k-Anonymity for Electronic Health Records: A Topological Framework 19th

Dynamic k-Anonymity for Electronic Health Records: A Topolo...

引用

19th International Workshop on data Privacy Management, DPM 2024, 8th International Workshop on Cryptocurrencies and Blockchain Technology, CBT 2024 and 10th Workshop on the Security of Industrial Control Systems and of Cyber-Physical Systems, CyberICPS 2024 which were held in conjunction with the 29th European Symposium on Research in computer Security, ESORICS 2024

作者： Swaminathan, Arjhun Akgün, Mete Medical Data Privacy and Privacy Preserving Machine Learning Department of Computer Science University of Tübingen Tübingen Germany Institute for Bioinformatics and Medical Informatics Tübingen Germany

ISBN: (纸本)9783031823480

With the rapid digitization of Electronic Health Records (EHRs), fast and adaptive data anonymization methods have become increasingly important. While tools from topological data analysis (TDA) have been proposed to anonymize static datasets—allowing the creation of multiple generalizations for different anonymization needs from a single computation—the application to dynamic datasets remains unexplored. To address this, our work adapts existing methodologies to the dynamic setting. We develop an improved version of weighted persistence barcodes that track higher-dimensional holes in data, allowing us to edit persistence information on the fly. Additionally, we introduce filtration trimming, a novel technique designed to update persistence data quickly with minimal computing effort when data is added. Our work represents a significant advancement in healthcare data privacy, offering a refined approach to protecting highly sensitive and evolving patient data through dynamic k-anonymity. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

Decoding Bias in #Climatechange Tweets: A Neural Language Style Transfer Approach

Decoding Bias in #Climatechange Tweets: A Neural Language St...

引用

2024 International Conference on Artificial Intelligence and Quantum Computation-Based Sensor Applications, ICAIQSA 2024

作者： George, Elizabeth Leah Parthasarathy, Subashini Avinashilingam Institute for Home Science and Higher Education for Women Department of Computer Science Coimbatore India Avinashilingam Institute for Home Science and Higher Education for Women Centre for Machine Learning and Intelligence Department of Computer Science Coimbatore India

ISBN: (纸本)9798331517953

In response to the increase of biased and misleading opinions on social media regarding climate change, we present the Neural Language Style Transfer Bias (NLST Bias) framework - an AI-driven solution to identify and mitigate bias in climate-related tweets. Our approach analyses a curated dataset of climate change tweets collected from Twitter, leveraging spaCy for text pre-processing and BERT-based sentence embeddings to extract semantic features. Evaluation metrics with classifier models showed significant improvements, with CNN-LSTM achieving the highest F1 score of 78%. Most classifiers show marked improvements across accuracy, precision, recall, and F1 score with NLST, highlighting its effectiveness on this dataset. NLST Bias employs DistilBERT to classify tweets as biased or unbiased, with the Named Entity Recognition (NER) model trained on RobERTa transformers to identify word-level bias indicators. The framework's core component is a transformer-based style transfer model (e.g., GPT-2) that modifies the stylistic tone of biased text while preserving its semantic content. This transfer uses a style-specific decoder within the transformer architecture, ensuring the generated text maintains grammatical accuracy and contextual relevance. The resulting output exhibits tone, phrasing, and word choice adjustments to foster a more neutral style. Results indicate a significant reduction in bias without compromising text fluency or naturalness, verified through statistical metrics and human evaluations. Furthermore, Explainable AI techniques are integrated to enhance transparency in model decisions. Our findings underscore NLST Bias's efficacy in countering climate misinformation and promoting balanced discussions on social media, highlighting its potential as a tool for responsible AI in digital discourse. © 2024 IEEE.

关键词： Tweets

来源：评论

学校读者我要写书评

暂无评论

Enhanced Steganalysis for Color Images Using Curvelet Features and Support Vector machine

引用

computers, Materials & Continua 2024年第1期78卷 1311-1328页

作者： Arslan Akram Imran Khan Javed Rashid Mubbashar Saddique Muhammad Idrees Yazeed Yasin Ghadi Abdulmohsen Algarni Faculty of Computer Science and Information Technology Department of Computer ScienceSuperior UniversityLahore54000Pakistan Machine Learning Code(MLC)Lab Meharban HouseHouse Number 209Zafar ColonyOkara56300Pakistan Information Technology Services University of OkaraOkara56310Pakistan Department of Computer Science&Engineering University of Engineering&Technology LahoreNarowal CampusNarowal51601Pakistan Department of Computer Science and Software Engineering Al Ain UniversityAl Ain11671Abu Dhabi Department of Computer Science King Khalid UniversityAbha61421Saudi Arabia

Algorithms for steganography are methods of hiding data transfers in media *** machine learning architectures have been presented recently to improve stego image identification performance by using spatial information,and these methods have made it feasible to handle a wide range of problems associated with image *** with little information or low payload are used by information embedding methods,but the goal of all contemporary research is to employ high-payload images for *** address the need for both low-and high-payload images,this work provides a machine-learning approach to steganography image classification that uses Curvelet transformation to efficiently extract characteristics from both type of *** Vector machine(SVM),a commonplace classification technique,has been employed to determine whether the image is a stego or *** Wavelet Obtained Weights(WOW),Spatial Universal Wavelet Relative Distortion(S-UNIWARD),Highly Undetectable Steganography(HUGO),and Minimizing the Power of Optimal Detector(MiPOD)steganography techniques are used in a variety of experimental scenarios to evaluate the performance of the *** WOW at several payloads,the proposed approach proves its classification accuracy of 98.60%.It exhibits its superiority over SOTA methods.

关键词： Curvelets fast fourier transformation support vector machine high pass filters steganography

来源：评论

学校读者我要写书评

暂无评论

Secure Healthcare Monitoring: Algorithms and Techniques for Heart Disease Detection and data Privacy 1

Secure Healthcare Monitoring: Algorithms and Techniques for ...

引用

1st International Conference on Intelligent Systems and Computational Networks, ICISCN 2025

作者： Poonam Mallapur, Sujata V. Sharnbasva University Department of Computer Science and Engineering Kalaburagi India Sharnbasva University Department of Artificial Intelligence & Machine Learning Kalaburagi India

ISBN: (纸本)9798331529246

This survey offers the review of Healthcare Monitoring Systems (HMS) and Privacy Preservation (PP) approaches. The main objective is based on the detection of heart disease and maintain the security for patient data. The machine learning models such as SVM, decision trees, and neural networks for heart disease detection and encryption approaches such as AES, ECC, and homomorphic encryption for securing the data are the major topics that reviewed in this survey. Moreover, the survey highlights the performance evaluation for heart disease prediction and the datasets used for heart disease. It also ensures the privacy to tackle the challenges in real time monitoring. A higher accuracy of 99.8 was reached by FHE-based encryption method. Moreover, it discusses the research gaps like scalability, efficiency, and regulatory compliance which motivates the need of novel privacy-preserving techniques for healthcare systems, which is highlighted in the future scope. © 2025 IEEE.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

引用

2023 International Conference on Artificial Intelligence science and Applications in Industry and Society, CAISAIS 2023

作者： Ghanem, Sahar I. Computer Science and Artificial Intelligence Faculty Artificial Intelligence and Machine Learning Department Egypt

ISBN: (纸本)9798350314786

In an attempt to facilitate the way students, succeed academically, academic advising is an integral part of the educational system. However, it has frequently been dependent on time-consuming manual processes and personalized interaction, which are not scalable. machine learning and deep learning have the potential to completely transform advising by automating parts of process and giving students distinctive guidance. This paper explores the relationship between student academic performance and correlated features using deep learning methodology in academic advising systems. The model was trained and tested with the Kaggle dataset, and the results show high accuracy, specificity, and true positive rate values compared to naive bayes and random forest methodologies. The paper also addresses potential challenges and limitations of deep learning in academic advising, such as data privacy concerns, algorithmic bias, and the need for human intervention. It concludes by emphasizing the importance of incorporating ethical considerations and human expertise in the development and deployment of deep learning models in academic advising. Overall, the paper highlights the potential of deep learning algorithms to transform academic advising and improve student success. © 2023 IEEE.

关键词： Students

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：