检索结果-内蒙古大学图书馆

Proceedings of the 37th International Conference on Neural Information Processing Systems

作者： Yubin Shi Yixuan Chen Mingzhi Dong Xiaochen Yang Dongsheng Li Yujiang Wang Robert Dick Qin Lv Yingying Zhao Fan Yang Tun Lu Ning Gu Li Shang China and Shanghai Key Laboratory of Data Science School of Computer Science Fudan University School of Mathematics Statistics University of Glasgow Microsoft Research Asia Shanghai China Department of Engineering Science University of Oxford Department of Electrical Engineering and Computer Science University of Michigan Department of Computer Science University of Colorado Boulder School of Microelectronics Fudan University

Despite their prevalence in deep-learning communities, over-parameterized models convey high demands of computational costs for proper training. This work studies the fine-grained, modular-level learning dynamics of over-parameterized models to attain a more efficient and fruitful training strategy. Empirical evidence reveals that when scaling down into network modules, such as heads in self-attention models, we can observe varying learning patterns implicitly associated with each module's trainability. To describe such modular-level learning capabilities, we introduce a novel concept dubbed modular neural tangent kernel (mNTK), and we demonstrate that the quality of a module's learning is tightly associated with its mNTK's principal eigenvalue λmax. A large λmax indicates that the module learns features with better convergence, while those miniature ones may impact generalization negatively. Inspired by the discovery, we propose a novel training strategy termed Modular Adaptive Training (MAT) to update those modules with their λmax exceeding a dynamic threshold selectively, concentrating the model on learning common features and ignoring those superfluous ones. Unlike most existing training schemes with a complete BP cycle across all network modules, MAT can significantly save computations by its partially-updating strategy and can further improve performance. Experiments show that MAT nearly halves the computational cost of model training and outperforms the accuracy of baselines.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Limiting behavior of bilinear forms for the resolvent of sample covariance matrices under elliptical distribution with applications

arXiv

引用

arXiv 2023年

作者： Yin, Yanqing Zhou, Wang School of Mathematics and Statistics Chongqing University China Department of Statistics and Data Science National University of Singapore Singapore

In this paper, we introduce a joint central limit theorem (CLT) for specific bilinear forms, encompassing the resolvent of the sample covariance matrix under an elliptical distribution. Through an exhaustive exploration of our theoretical findings, we unveil a phase transition in the limiting parameters that relies on the moments of the random radius in our derived CLT. Subsequently, we employ the established CLT to address two statistical challenges under elliptical distribution. The first task involves deriving the CLT for eigenvector statistics of the sample covariance matrix. The second task aims to ascertain the limiting properties of the spiked sample eigenvalues under a general spiked model. As a byproduct, we discover that the eigenmatrix of the sample covariance matrix under a light-tailed elliptical distribution satisfies the necessary conditions for asymptotic Haar, thereby extending the Haar conjecture to broader *** Codes 62H15, 62B20 © 2023, CC BY.

关键词： Covariance matrix

来源：评论

学校读者我要写书评

暂无评论

Multi to binary class size based imbalance handling technique in wireless sensor networks

Multi to binary class size based imbalance handling techniqu...

引用

作者： Singh, Neha Virmani, Deepali Dhiman, Gaurav Vimal, S. University School of Information Communication and Technology Guru Gobind Singh Indraprastha University Delhi Dwarka110086 India Department of Computer Science Engineering VIPS-TC School of Engineering and Technology Delhi110034 India Department of Electrical and Computer Engineering Lebanese American University Byblos Lebanon Department of Computer Science and Engineering University Centre for Research and Development Chandigarh University Mohali Gharuan140413 India Department of Computer Science and Engineering Graphic Era Deemed to be University Dehradun248002 India Division of Research and Development Lovely Professional University Phagwara144402 India Chitkara University Institute of Engineering and Technology Chitkara University Punjab 140601 India Department of Computer Science Government Bikram College of Commerce Punjab Patiala147001 India Department of Artificial Intelligence and Data Science Ramco Institute of Technology Tamil Nadu Rajapalayam626117 India

Wireless sensor network (WSN) has entered various disciplines including healthcare, banking, transportation, ocean and wildlife monitoring, earthquake monitoring, and numerous military applications. Now days, there is escalation in size of data which makes it unfeasible to analyse it with accuracy. There are numerous problems that are faced when detecting a pattern between structured and unstructured data which are unworkable by humans, so to make computation fast, easy and accurate, Machine Learning came in existence. Machine Learning is extensively used in WSNs. To make a machine learn, a training dataset is required and output is predicted by testing the dataset. A dataset in WSN has multi-class in its dependent variable. This multiclass classification causes class imbalance problem. The paper proposes multi to binary class size based imbalance handling technique in wireless sensor networks (MBSCIH) technique to solve class imbalance problem in multi-class classification. The technique MBSCIH converts multi-class classification into binary-class classification. The proposed technique MBSCIH is applied on WSN-DS, NSL-KDD and KDD-cup 99 datasets and is tested with five major machine learning algorithm: Naive Bayes, Random Forest, decision tree, support vector machine (SVM) and k-nearest neighbour (KNN). The test method used for testing is 10-fold cross validation. Results infer that the proposed method increases the existing efficiency by 15.13%, 0.28%, 0.01%, 0.01%, 0.12% for Naïve Bayes, Decision Tree, SVM, Random Forest and KNN Algorithms respectively for KDD Cup99 dataset. The proposed method increases the efficiency by 43.1%, 2.79%, 0.01%, 0.53% for Naïve Bayes, decision tree, SVM and KNN algorithms respectively for NSL-KDD dataset. Also, the proposed method increases the efficiency by 0.13%, 0.08%, 0.29%, 0.29% for Naïve Bayes, decision tree, SVM and KNN algorithms respectively for WSN-DS *** the experimental analysis proves that solving multiclas

关键词： Wireless sensor networks

来源：评论

学校读者我要写书评

暂无评论

EEG-Driven Machine Learning Framework for Student Cognitive Load Classification

EEG-Driven Machine Learning Framework for Student Cognitive ...

引用

Smart Systems for applications in Electrical sciences (ICSSES), International Conference on

作者： Varunkumar MS Siddharth Gupta Pramod TC Department of Artificial Intelligence and Data Science Siddaganga Institute of Technology Tumkur Karnataka India Department of Computer Science & Engineering Siddaganga Institute of Technology Tumkur Karnataka India

ISBN: (数字)9798331507671

ISBN: (纸本)9798331507688

Understanding differences in learning abilities is essential for improving education. This paper presents a framework to differentiate between learners using EEG data. The framework identifies critical factors that influence learning experiences by applying data-driven approaches and statistical methods. It includes analyses such as band power, entropy calculation, and coherence measurement across EEG channels. The extracted features are used to train a classifier and to identify cognitive behavior patterns. The framework's accuracy is verified using cross-validation, classification reports, and confusion matrices. The experimental analysis shows major improvements in cognitive pattern understanding and identification. This framework provides educators with actionable insights to design personalized learning strategies. By enhancing the understanding of individual learning dynamics, the framework has the potential to optimize educational outcomes across diverse learner profiles.

关键词： Support vector machines Accuracy Education Machine learning Coherence Feature extraction Brain modeling Cognitive load Electroencephalography Entropy

来源：评论

学校读者我要写书评

暂无评论

DMGNN: Cancer driver gene identification based on mix-moment graph neural network

DMGNN: Cancer driver gene identification based on mix-moment...

引用

2024 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2024

作者： Chen, Bolin Li, Ziyuan Li, Haodong Li, Junming Kang, Ruiming Shang, Xuequn Bian, Jun Northwestern Polytechnical University School of Computer Science Xian China Northwestern Polytechnical University Key Laboratory of Big Data Storage and Management School of Computer Science China Information Technology Co. Ltd Hangzhou China Xian Jiaotong University Affiliated Childrens Hospital Xian Childrens Hospital Department of General Surgery Xian China

ISBN: (纸本)9798350386226

Cancer development is closely linked to the accumulation of mutations in driver genes. Therefore, identification of driver genes is crucial for understanding the molecular basis of cancer. In various types of methods, approches based on Graph Neural Networks (GNNs) framework are one of the most effective tools to identify driver genes, which can also combine biological networks and multi-omics data to further improve the identification accuracy. However, many GNN frameworks often utilize single-order moment to get neighbourhood information for message passing, which ignores the rich distribution information and gene features of neighbouring genes. Besides, when using GNNs model, it is necessary to stack the hidden layers, which can often lead to both over-smoothing and network degradation problems. To overcome these issues, a new framework called DMGNN was proposed for identifying driver genes. To get rich distribution information and gene features, a mix-moment embedding and attention-based feature selection were utilized among neighbouring genes. To solve the problem posed by hidden layer stacking in the GNN model, a deepwalk method was used to learn remote gene effects for the over-smoothing problem, and the resNet-based hidden layer aggregation was employed to mitigate the network degradation issue. Experimental results demonstrate that the proposed model outperforms many existing methods for identifying cancer driver genes, where the AUROC value was achieved at 0.856 in STRING dataset with at least 2 percentage improvement comparing with other models. The DMGNN is freely available via https://***/lavendar682/DMGNN. © 2024 IEEE.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Searching for local associations while controlling the false discovery rate

arXiv

引用

arXiv 2024年

作者： Gablenz, Paula Sesia, Matteo Sun, Tianshu Sabatti, Chiara Department of Statistics Stanford University CA United States Departments of Data Sciences and Operations and of Computer Science University of Southern California CA United States Cheung Kong Graduate School of Business Beijing China Departments of Biomedical Data Science Statistics Stanford University CA United States

We introduce local conditional hypotheses that express how the relation between explanatory variables and outcomes changes across different contexts, described by covariates. By expanding upon the model-X knockoff filter, we show how to adaptively discover these local associations, all while controlling the false discovery rate. Our enhanced inferences can help explain sample heterogeneity and uncover interactions, making better use of the capabilities offered by modern machine learning models. Specifically, our method is able to leverage any model for the identification of data-driven hypotheses pertaining to different contexts. Then, it rigorously test these hypotheses without succumbing to selection bias. Importantly, our approach is efficient and does not require sample splitting. We demonstrate the effectiveness of our method through numerical experiments and by studying the genetic architecture of Waist-Hip-Ratio across different sexes in the UKBiobank. © 2024, CC BY-NC-ND.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Cluster analysis: application of K-means algorithm to explore the sea surface temperature and salinity patterns

Cluster analysis: application of K-means algorithm to explor...

引用

2023 International Conference on computer Application and Information Security, ICCAIS 2023

作者： Ding, Keyan Liu, Zihan Wang, Mo School of Computing and Data Science Xiamen University Malaysia Jalan Sunsuria Bandar Sunsuria Selangor Darul Ehsan Sepang43900 Malaysia School of Mathematics and Statistics The University of Sheffield Sheffield United Kingdom Department of Statistical Science University College London LondonWC1E 6BT United Kingdom

ISBN: (纸本)9781510675032

This paper investigates how the sea surface temperature (SST) and salinity (SSS) change with the seasons based on K-means clustering analysis. For index k, the elbow method is used to estimate the value. Then this paper analyzes the clustering of SSS and SST in a specific region that contains both characteristic SST and SSS features and compare the clustering figures from each season. Meanwhile, by comparing K-means cluster results computed from raw data to those obtained for classified histograms, we find that it is ideal to combine these two types of data to do an analysis. To illustrate our methodology, we use the Silhouette Coefficient to find the goodness of our K-means cluster, and the calculated Silhouette Coefficient equals 0.8, which illustrates our model has relatively high quality and accuracy. Lastly, we apply other clusters, such as DBSCAN and Gaussian Mixture Model to the same data, but the results indicate the difference between our data is too weak to be clustered by these two methods. © 2024 SPIE.

关键词： Surface temperature

来源：评论

学校读者我要写书评

暂无评论

Human-computer Interface Challenges in Mobile data Mining

Human-Computer Interface Challenges in Mobile Data Mining

引用

2023 International Conference on Sustainable Computing and Smart Systems, ICSCSS 2023

作者： Padmanaban, K. Sathiya, A. Jeevitha, K. Rosaline, R. Anto Arockia Koneru Lakshmaiah Education Foundation Department of Computer Science and Engineering Andhra Pradesh Vijayawada India Sri Sairam Institute of Technology Department of Artificial Intelligence and Data Science Tamil Nadu Chennai India R.M.K. Engineering College Department of Electronics and Communication Engineering Tamil Nadu Chennai India Srm Institute of Science and Technology Department of Computing Technologies Tamil Nadu Chennai India

ISBN: (纸本)9798350333602

The continuous advancement of computer software, hardware, and artificial intelligence, has allowed for tremendous progress in the development of human-machine communication systems in recent years. The development of such methods provides more accurate results than humans. The core of the human-computer interface is the Natural Language Understanding (NLU) component. The success of Human-computer Interaction (HCI) and the time it takes to accomplish tasks are both directly affected by the results of this module's semantic comprehension. In this research study, a neural network and HCI-based engine is developed for understanding and interacting with natural languages. data from the appropriate literature is used to comprehend the necessary understanding of HCI. Next, the interactive engine is used to develop a model of HCI based on natural language comprehension using a neural network, and also the model's underlying neural network reasoning structure is assessed. © 2023 IEEE.

关键词： Engines

来源：评论

学校读者我要写书评

暂无评论

Predicting Malware Classification and Family using Machine Learning: A Cuckoo Environment Approach with Automated Feature Selection 2

Predicting Malware Classification and Family using Machine L...

引用

2nd International Conference on Machine Learning and data Engineering, ICMLDE 2023

作者： Niveditha, S. Prianka, R.R. Sathya, K. Shreyanth, S. Subramani, Nandhagopal Deivasigamani, Balakrishnan Karthikeyan, S. Department of Biotechnology Rajalakshmi Engineering College Thandalam Tamil Nadu Chennai602105 India Department of AIML New Horizon College of Engineering Karnataka Bengaluru560103 India School of Computer Science and Engineering VIT Tamilnadu Vellore632014 India MTech in Data Science and Engineering Birla Institute of Technology and Science Rajasthan Pilani333031 India KPR Institute of Engineering and Technology Tamil Nadu Coimbatore641407 India Department of Information Technology KSR Institute for Engineering and Technology India

Malware's increasing menace in the digital realm needs the development of powerful detection and classification systems. This study presents a unique method for predicting malware category and family using machine learning, leveraging the Cuckoo environment and automated feature selection. To undertake an exhaustive examination of malware activity, we combined the Cuckoo environment with cutting-edge machine learning methods, such as the k-Nearest Neighbors (KNN) technique. The dynamic analysis of malware samples within the Cuckoo environment, which captured their interactions with the execution environment, provided remarkable insights into their malicious behaviors. The Boruta method aided automated feature selection, improving the feature set and optimizing model performance. A comparison with existing models yielded striking findings. Notably, Test Case 3 surpassed earlier cases by incorporating Automated Feature Selection + Cross-Validation. In identifying phishing attempts, it displayed outstanding specificity (90%), precision (93%), recall (96%), and an impressive F1-Score (92%), indicating its proficiency in reliably recognizing this frequent threat. Test Case 3 showed considerable increases as well, with a surprising 7.0% increase in precision and a notable 14.6% increase in recall when compared to Test Case 1. Furthermore, in the field of ransomware detection, Test Case 4, which solely focused on Automated Feature Selection, obtained excellent results, with 91% specificity, 92% accuracy, and a recall rate of 93%. These developments highlight the importance of the Boruta algorithm in optimizing the model's performance. © 2024 Elsevier B.V.. All rights reserved.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

MARKOV CHAIN ENTROPY GAMES AND THE GEOMETRY OF THEIR NASH EQUILIBRIA

arXiv

引用

arXiv 2023年

作者： Choi, Michael C.H. Wolfer, Geoffrey DEPARTMENT OF STATISTICS AND DATA SCIENCE AND YALE-NUS COLLEGE NATIONAL UNIVERSITY OF SINGAPORE Singapore WASEDA UNIVERSITY CENTER FOR DATA SCIENCE Tokyo Japan

Consider the following two-person mixed strategy game of a probabilist against Nature with respect to the parameters (f, B, π), where f is a convex function satisfying certain regularity conditions, B is either the set {Li}ni=1 or its convex hull with each Li being a Markov infinitesimal generator on a finite state space X and π is a given positive discrete distribution on X. The probabilist chooses a prior measure µ within the set of probability measures on B denoted by P(B) and picks a L ∈ B at random according to µ, whereas Nature follows a pure strategy to select M ∈ L(π), the set of π-reversible Markov generators on X. Nature pays an amount Df(M||L), the f-divergence from L to M, to the probabilist. We prove that a mixed strategy Nash equilibrium always exists, as well as a minimax result of the form (Fomular Presented) This also contrasts with the pure strategy version of the game where we show a Nash equilibrium may not exist. To find approximately a mixed strategy Nash equilibrium, we propose and develop a simple projected subgradient algorithm that provably converges with a rate of O(1/√t), where t is the number of iterations. In addition, we elucidate the relationships of Nash equilibrium with other seemingly disparate notions such as weighted information centroid, Chebyshev center and Bayes risk. This article generalizes the two-person game of a statistician against Nature developed in [D. Haussler (1997). A general minimax result for relative entropy. IEEE Trans. Inform. Theory 43: 1276–1280;D. Haussler and M. Opper (1997). Mutual information, metric entropy and cumulative relative entropy risk. Ann. Statist. 25(6): 2451-2492;A.A. Gushchin and D.A. Zhdanov (2006). A minimax result for f-divergences. In: From Stochastic Calculus to Mathematical Finance. Springer, Berlin, Heidelberg.], and highlights the powerful interplay and synergy between modern Markov chains theory and geometry, information theory, game theory, optimization and mathematical statistics

关键词： Markov chains

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：