检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Abdelmalak, Ibram Madhusudhanan, Kiran Choi, Jungmin Stubbemann, Maximilian Schmidt-Thieme, Lars Information Science and Machine Learning Lab VWFS Data Analytics Research Center University of Hildesheim Niedersachsen Hildesheim Germany

Time-series forecasting research has converged to a small set of datasets and a standardized collection of evaluation scenarios. Such a standardization is to a specific extent needed for comparable research. However, the underlying assumption is, that the considered setting is a representative for the problem as a whole. In this paper, we challenge this assumption and show that the current scenario gives a strongly biased perspective on the state of time-series forecasting research. To be more detailed, we show that the current evaluation scenario is heavily biased by the simplicity of the current datasets. We furthermore emphasize, that when the lookback-window is properly tuned, current models usually do not need any information flow across channels. However, when using more complex benchmark data, the situation changes: Here, modeling channel-interactions in a sophisticated manner indeed enhances performances. Furthermore, in this complex evaluation scenario, Crossformer, a method regularly neglected as an important baseline, is the SOTA method for time series forecasting. Based on this, we present the Fast Channel-dependent Transformer (FaCT), a simplified version of Crossformer which closes the runtime gap between Crossformer and TimeMixer, leading to an efficient model for complex forecasting datasets. © 2025, CC BY.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Robust and Efficient Writer-Independent IMU-Based Handwriting Recognization

arXiv

引用

arXiv 2025年

作者： Li, Jindong Hamann, Tim Barth, Jens Kaempf, Peter Zanca, Dario Eskofier, Bjoern Machine Learning and Data Analytics Lab Friedrich-Alexander-Universität Erlangen-Nürnberg Germany STABILO International GmbH Germany

Online handwriting recognition (HWR) using data from inertial measurement units (IMUs) remains challenging due to variations in writing styles and the limited availability of high-quality annotated datasets. Traditional models often struggle to recognize handwriting from unseen writers, making writer-independent (WI) recognition a crucial but difficult problem. This paper presents an HWR model with an encoder-decoder structure for IMU data, featuring a CNN-based encoder for feature extraction and a BiLSTM decoder for sequence modeling, which supports inputs of varying lengths. Our approach demonstrates strong robustness and data efficiency, outperforming existing methods on WI datasets, including the WI split of the OnHW dataset and our own dataset. Extensive evaluations show that our model maintains high accuracy across different age groups and writing conditions while effectively learning from limited data. Through comprehensive ablation studies, we analyze key design choices, achieving a balance between accuracy and efficiency. These findings contribute to the development of more adaptable and scalable HWR systems for real-world applications. The code is available at https://***/jindongli24/REWI. Copyright © 2025, The Authors. All rights reserved.

关键词： Units of measurement

来源：评论

学校读者我要写书评

暂无评论

Noise filtering approach to improve handwritten digit recognition using customized CNN for Cerebral Palsy individuals

引用

The European Physical Journal Special Topics 2025年 1-19页

作者： Muthureka, K. Srinivasulu Reddy, U. Janet, B. Machine Learning and Data Analytics Lab Department of Computer Applications National Institute of Technology-Tiruchirappalli Tiruchirappalli India Centre of Excellence (CoE) in Artificial Intelligence National Institute of Technology-Tiruchirappalli Tiruchirappalli India Information Processing Lab Department of Computer Applications Tiruchirappalli India

Automatic recognition of handwritten digits in people with Cerebral Palsy (CP) is a serious challenge that requires advancements in data preparation for improved predictive accuracy. The disorganized digit pixel patterns caused by CP people’ innate motor coordination problems result in the classification errors due to the noise and outlier. Despite the fact that many handwritten digit databases for normal people are available, none expressly cater to those with shaky hands/CP dataset. This highlights the importance of an efficient pre-processing method for mitigating variances in CP individuals’ digit pixel patterns. A total of 43,000 images from 43 CP people were collected and organized into 10 different folders for data labeling. The authors used the K-Means Based Noise Removal (KMNR) approach to remove noise and outliers from a CP and tested it on a customized CNN. The KMNR model achieved an accuracy of 98.86%, compared to 93.54% without it. The F1-score values for state-of-the-art models without KMNR ranged from 85.00% to 97.00%, but after integrating KMNR, they increased to 92.00% to 98.00%. The precision values ranged from 85.00% to 97.00% in the absence of KMNR, but increased to 91.00% to 98.00% in the presence of KMNR. The recall values ranged from 85.00% to 97.00% without KMNR, but increased to 87.00% to 98.00% with KMNR.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Problems and Prospectives of Big data Storage and Processing Standartization

Problems and Prospectives of Big Data Storage and Processing...

引用

2019 International Multi-Conference on Engineering, Computer and Information Sciences, SIBIRCON 2019

作者： Pavlovskiy, Evgeniy N. Stream Data Analytics and Machine Learning Lab Novosibirsk State University Novosibirsk Russia

ISBN: (纸本)9781728144016

In the paper, we analyze the problem of standardization in the domain of storage and processing of big data in the application to the Internet of things. We highlight the underlying problems of big data;analyze the scientific, technological and economic barriers to the development of big data. We offer perspective research directions in the domain of development of standards of data representation. © 2019 IEEE.

关键词： data structures

来源：评论

学校读者我要写书评

暂无评论

Binary Brain Tumor Classification With Semantic Features Using Convolutional Neural Network

Binary Brain Tumor Classification With Semantic Features Usi...

引用

2022 Ural Symposium on Biomedical Engineering, Radioelectronics and Information Technology, USBEREIT 2022

作者： Khue, Luu Minh Sao Pavlovskiy, Evgeniy Novosibirsk State University Stream Data Analytics and Machine Learning Laboratory Novosibirsk Russia

ISBN: (数字)9781665460927

ISBN: (纸本)9781665460927

In this study, we provide segmentations of brain tumors as semantic features to a simple convolutional neural network (CNN) to improve the classification results. The Siberian Brain Tumor dataset (SBT) of 1452 magnetic resonance (MR) images of Russian people's brains is used for training, validation, and testing. We preprocess MR images by removing the swelling region surrounding a brain tumor (edema) from the segmentation and eliminating slices that contain less than 163 pixels of tumor. The binary classifier is a simple network of four convolutional layers for feature extractions and two linear layers for classification. The network is trained and validated with 5-fold cross-validation. We train another CNN with similar configuration on images without provided segmentation and compare the testing results. The two networks are evaluated with four metrics: accuracy, sensitivity, specificity, and F1 scores. The classifier trained with segmentation achieve the highest accuracy score of 0.92, sensitivity of 0.934, specificity of 0.91, and F1 score of 0.90 using ensemble approach. © 2022 IEEE.

关键词： Magnetic resonance

来源：评论

学校读者我要写书评

暂无评论

Improving Brain Tumor Multiclass Classification With Semantic Features

Improving Brain Tumor Multiclass Classification With Semanti...

引用

2022 Ural-Siberian Conference on Computational Technologies in Cognitive Science, Genomics and Biomedicine, CSGB 2022

作者： Sao Khue, Luu Minh Pavlovskiy, Evgeniy Novosibirsk State University Stream Data Analytics and Machine Learning Laboratory Novosibirsk Russia

ISBN: (数字)9781665452885

ISBN: (纸本)9781665452885

Histopathological examination of biopsy tissues is still utilized to diagnose and classify brain cancers today. The current approach is inconvenient, time-consuming, and prone to human mistake. These disadvantages emphasize the significance of establishing a fully automated deep learning-based system for classifying brain tumors. In this paper, we suggest an approach to improve the classification for four types of brain tumors by providing the classifier with segmentation as semantic features. 1,452 multi model magnetic resonance images from the Siberian Brain Tumor dataset (SBT) are used for training, validation, and testing. The training and validation are implemented with our experimental simple convolutional neural network and a pre-trained VGG16. Best performed models are selected and tested on both SBT and the Brain Tumor Segmentation Challenge 2020 dataset (BraTS). The models with segmentation outperform all models without segmentation on the same dataset. We also found that, compare to a general purposed network such as VGG16, a simple convolutional neural network trained on a specific task have better generalization when tested with a public dataset. © 2022 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

System Architecture for Deep Packet Inspection in High-speed Networks

System Architecture for Deep Packet Inspection in High-speed...

引用

Siberian Symposium on data Science and Engineering (SSDSE)

作者： Khazankin, Grigory R. Komarov, Sergey Kovalev, Danila Barsegyan, Artur Likhachev, Alexander Novosibirsk State Univ Stream Data Analyt & Machine Learning Lab Novosibirsk Russia

ISBN: (纸本)9781538615935

To solve the problems associated with large data volume real-time processing, heterogeneous systems using various computing devices are increasingly used. The characteristic of solving this class of problems is related to the fact that there are two directions for improving methods of real-time data analysis: the first is the development of algorithms and approaches to analysis, and the second is the development of hardware and software. This article reviews the main approaches to the architecture of a hardware-software solution for traffic capture and deep packet inspection (DPI) in data transmission networks with a bandwidth of 80 Gbit/s and higher. At the moment there are software and hardware tools that allow designing the architecture of capture system and deep packet inspection: Using only the central processing unit (CPU);Using only the graphics processing unit (GPU);Using the central processing unit and graphics processing unit simultaneously (CPU + GPU). In this paper, we consider these key approaches. Also attention is paid to both hardware and software requirements for the architecture of solutions. Pain points and remedies are described.

关键词： Napatech Mellanox DPDK PeerDirect DPI Nvidia CUDA GPUDirect Zero-copy DMA

来源：评论

学校读者我要写书评

暂无评论

The EmpkinS-EKSpression Reappraisal Training Augmented With Kinesthesia in Depression: One-Armed Feasibility Study

引用

JMIR Formative Research 2025年 9卷 e65357页

作者： Keinert, Marie Schindler-Gmelch, Lena Rupp, Lydia Helene Sadeghi, Misha Richer, Robert Capito, Klara Eskofier, Bjoern M. Berking, Matthias Department of Clinical Psychology and Psychotherapy Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) Erlangen Germany Machine Learning and Data Analytics Lab Department of Artificial Intelligence in Biomedical Engineering (AIBE) Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) Erlangen Germany Translational Digital Health Group Institute of AI for Health Helmholtz Zentrum Muenchen Neuherberg Germany

Background: Harboring dysfunctional depressogenic cognitions contributes to the development and maintenance of depression. A central goal of cognitive behavioral therapy (CBT) for depression is to invalidate such cognitions via cognitive reappraisal (CR). However, relatively low remission rates and high dropout rates in CBT demonstrate the need for further improvement. Potentially, the effects of CBT could be enhanced by addressing not only dysfunctional depressogenic cognitions but also body states associated with depression. This may be done, for example, by systematically pairing the invalidation of depressogenic cognitions with the performance of antidepressive kinesthesia. Objective: This study aimed to examine the feasibility and clinical potential of a smartphone-based cognitive restructuring task that required users to deliberately perform antidepressive kinesthesia in conjunction with the rejection of depressogenic statements and the affirmation of antidepressive statements. This feasibility study was conducted as a precursor to a large-scale randomized controlled trial. Methods: In total, 10 healthy participants engaged in a single 90-120-minute session of smartphone-based CR training. During the training, they completed 2 phases in which they were required to reject 20 depressogenic and affirm 20 antidepressive statements, respectively. Diagnostic assessments were conducted 1 week (T1) before and directly prior (T2) to the training, and again directly posttraining (T3) and at a 2-week follow-up posttraining (T4). Feasibility outcomes assessed at T3 included intervention safety recorded by study therapists, compliance, technical feasibility, usability assessed using the Short Version of the User Experience Questionnaire (UEQ-S), and acceptability assessed using the UEQ-S and self-developed items. Preliminary clinical potential was evaluated via single-item ratings of current depressed and positive mood assessed continuously during the training. Feasibility

关键词： cognitive reappraisal depression facial expression kinesthesia mobile phone smartphone-based intervention

来源：评论

学校读者我要写书评

暂无评论

Automated Pipeline for Regional Epicardial Adipose Tissue Distribution Analysis in the Left Atrium

Automated Pipeline for Regional Epicardial Adipose Tissue Di...

引用

15th International Workshop on Statistical Atlases and Computational Models of the Heart, STACOM 2024, Held in Conjunction with MICCAI 2024

作者： Lluch, Èric Castañeda, Eduardo Weiss, Maximilian E. R. Vizitiu, Anamaria Jacob, Athira Jami, Jitin Audigier, Chloé Meister, Felix Mihalef, Viorel Passerini, Tiziano Siemens Healthineers Digital Technologies and Innovation Erlangen Germany Pattern Recognition Lab Department of Computer Science Friedrich-Alexander University Erlangen-Nürnberg Erlangen Germany Siemens SRL Brasov Romania Siemens Healthineers Digital Technology and Innovation Princeton United States Machine-Learning and Data Analytics Friedrich-Alexander-Universität Erlangen-Nürnberg Erlangen Germany Advanced Clinical Imaging Technology Siemens Healthineers AG Lausanne Switzerland

ISBN: (纸本)9783031877551

Atrial fibrillation (AF) is the most common cardiac arrhythmia, affecting approximately 3% of the global population and rising to 11% in individuals over 80 years old. The distribution of epicardial adipose tissue (EAT) around the atria plays a significant role in the pathophysiology of AF, yet its regional analysis remains underexplored. This paper introduces an automated pipeline capable of segmenting 14 atrial regions from patient CT angiograms, a notable advancement over previous attempts to automate the creation of synthetic anatomical or physics-based twins. Our fully automatic pipeline requires no human supervision and provides the regional EAT quantification successfully on 91.9% of the analyzed patients for which the LA and its structures were automatically segmented (1797 out of 1954 patients), including the automatic segmentation of the aorta and pericardium to assess EAT volume around the left atrium. By applying principal component analysis of the regional EAT to analyze a separate dataset of 84 individuals who were clinically labeled, we identified a distinct group that contained all the patients with AF. This finding suggests that there is a significant link between the pattern of fat distribution and AF. This novel approach has the potential to enhance our understanding of AF and improve patient outcomes by enabling large-scale, regional fat analysis in an automated manner on all patients with accurate segmentation of left atrial structures. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Principal component analysis

来源：评论

学校读者我要写书评

暂无评论

Harnessing Ensemble machine learning Models for Timely Diagnosis of Breast Cancer Metastasis: A Case Study on CatBoost, XGBoost, and LGBM 25

Harnessing Ensemble Machine Learning Models for Timely Diagn...

引用

25th IEEE International Conference of Young Professionals in Electron Devices and Materials, EDM 2024

作者： Luu, Minh Sao Khue Banerjee, Santanu Pavlovskiy, Evgeniy N. Tuchinov, Bair N. Stream Data Analytics and Machine Learning Laboratory Novosibirsk State University Novosibirsk Russia Kharagpur West Bengal Kharagpur721302 India

ISBN: (纸本)9798350389234

This study employs three advanced gradient boosting machine learning algorithms to assess potential disparities in healthcare delivery. We specifically investigate which factors contribute to a patient's timely diagnosis of metastatic breast cancer using a public healthcare dataset. Our approach involves training and testing three separate models, as well as an ensemble model with automatically optimized weights. The models try to predict whether patients received a diagnosis of metastatic breast cancer within 90 days. Each model has different preprocessing and feature selection steps. The hyperparameter optimization is performed using the Optuna library in Python. Models are evaluated on the Kaggle platform, with our metrics indicating strong predictive performance;Categorical Boosting achieved an Area Under the Receiver Operating Characteristic Curve score of 0.813, Extreme Gradient Boosting reached 0.808, Light Gradient Boosting machine scored 0.805, and the ensemble model culminated at 0.808. Additionally, we analyze the set of features being used by all the best models and examine the impact of hyperparameters on the models' overall performance. © 2024 IEEE.

关键词： Diseases

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：