检索结果-内蒙古大学图书馆

Transfer Learning Using Variational Quantum Circuit 1

11th International Advanced computing Conference, IACC 2021

作者： Karur Mudugal Mathad, Rajashekharaiah Saurabh, Abhishek Mishra, Aditya Jain, Sambhav Kumar, Purushottam Vardaan Chickerur, Satyadhyan School of Computer Science and Engineering K L E Technological University Karnataka Hubballi580 031 India Centre for High Performance Computing K L E Technological University Karnataka Hubballi580 031 India

ISBN: (数字)9783030955021

ISBN: (纸本)9783030955014

Quantum computing and particularly quantum machine learning has witnessed a surge in the direction of fruitful research being done in recent years. Variational quantum algorithms have gained popularity due to their resilience to error, which is a common occurrence in the near term intermediate scale quantum (NISQ) hardware. In such an algorithm, an optimal function is mapped by a quantum circuit which is designed using weighted parameterized gates to be trained based on a cost function that is optimized using gradient descent. A Variational Quantum Classifier (VQC) is such an algorithm and is used for supervised learning to perform data classification. But since the quantum hardware available today is not powerful enough for a VQC to come of any practical use in the current setting, we implement a transfer learning pipeline in which a pre-trained image processing model is fine tuned using a dressed VQC;for a specific, binary classification task. Our aim was to study the working of a VQC and test it’s behaviour with respect to multiple hyperparameters such as circuit depth, number of qubits, etc. We trained our transfer learning-VQC model for 4 different datasets and for different hyperparameters and were able to get close to state of the art performance in most of our tests, solidifying the prospect of VQC as an efficient quantum-classical hybrid supervised learning algorithm. It should be noted that the study of VQC was chosen over that of other algorithms available for classification because recent advancements in the understanding of VQCs that show promise in favour of the algorithm. © 2022, Springer Nature Switzerland AG.

关键词： Learning algorithms

来源：评论

学校读者我要写书评

暂无评论

Online demand peak shaving with machine-learned advice in digital twins

Digital Twins and Applications

引用

Digital Twins and Applications 2024年第1期1卷 38-50页

作者： Minxi Feng Wei Li Boyu Qin Albert Y.Zomaya Centre for Distributed and High Performance Computing School of Computer ScienceThe University of SydneySydneyNew South WalesAustralia State Key Laboratory of Electrical Insulation and Power Equipment Department of Electrical EngineeringXi'an Jiaotong UniversityXi'anChina

As the use of physical instruments grows,control algorithms are being increasingly deployed to enhance efficiency and reliability through digital twin *** load management is central to energy systems within digital twins,which significantly impacts operational *** demand loads can lead to substantial monthly utility expenses without proper ***,a randomised online algorithm incorporating machine-learned insights is introduced to optimise battery operations and mitigate peak demand *** leverages limited-bit information from ma-chine learning models to inform its online decision-making process for cost-effective load *** provide theoretical evidence demonstrating that AMPAMOD maintains minimal advice complexity,has a linear computational cost,and achieves a bounded competitive *** trace-driven experiments with real-world household data reveal that AMPAMOD successfully reduces peak loads by over 90%,outperforming other benchmarks by at least 50%.These experimental findings align with our theoretical assertions,showcasing the effectiveness of AMPAMOD.

关键词： operations maintain battery

来源：评论

学校读者我要写书评

暂无评论

Towards Harmless Rawlsian Fairness Regardless of Demographic Prior

arXiv

引用

arXiv 2024年

作者： Wang, Xuanqian Li, Jing Tsang, Ivor W. Ong, Yew-Soon School of Computer Science and Engineering Beihang University China Institute of High-Performance Computing Agency for Science Technology and Research Singapore Centre for Frontier AI Research Agency for Science Technology and Research Singapore College of Computing and Data Science Nanyang Technological University Singapore

Due to privacy and security concerns, recent advancements in group fairness advocate for model training regardless of demographic information. However, most methods still require prior knowledge of demographics. In this study, we explore the potential for achieving fairness without compromising its utility when no prior demographics are provided to the training set, namely harmless Rawlsian fairness. We ascertain that such a fairness requirement with no prior demographic information essential promotes training losses to exhibit a Dirac delta distribution. To this end, we propose a simple but effective method named VFair to minimize the variance of training losses inside the optimal set of empirical losses. This problem is then optimized by a tailored dynamic update approach that operates in both loss and gradient dimensions, directing the model towards relatively fairer solutions while preserving its intact utility. Our experimental findings indicate that regression tasks, which are relatively unexplored from literature, can achieve significant fairness improvement through VFair regardless of any prior, whereas classification tasks usually do not because of their quantized utility measurements. The implementation of our method is publicly available at https://***/wxqpxw/VFair. © 2024, CC BY.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

Dual-Scale Enhanced and Cross-Generative Consistency Learning for Semi-Supervised Medical Image Segmentation

SSRN

引用

SSRN 2024年

作者： Gu, Yunqi Zhou, Tao Zhang, Yizhe Zhou, Yi He, Kelei Gong, Chen Zhu, Huafu PCA Lab The Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information Ministry of Education School of Computer Science and Engineering Nanjing University of Information Science and Technology Nanjing210094 China School of Computer Science and Engineering Southeast University Nanjing211189 China Medical School Nanjing University National Institute of Healthcare Data Science Nanjing210023 China Institute of High Performance Computing A*STAR Singapore

Medical image segmentation plays a crucial role in computer-aided diagnosis. However, existing methods heavily rely on fully supervised training, which requires a large amount of labeled data with time-consuming pixel-wise annotations. Moreover, accurately segmenting lesions poses challenges due to variations in shape, size, and location. To address these issues, we propose a novel Dual-scale Enhanced and Cross-generative consistency learning framework for semi-supervised medical image Segmentation (DEC-Seg). First, we propose a Cross-level Feature Aggregation (CFA) module that integrates cross-level adjacent layers to enhance the feature representation ability across different resolutions. To address scale variation, we present a scale-enhanced consistency constraint, which ensures consistency in the segmentation maps generated from the same input image at different scales. This constraint helps handle variations in lesion sizes and improves the robustness of the model. Furthermore, we propose a cross-generative consistency scheme, in which the original and perturbed images can be reconstructed using cross-segmentation maps. This consistency constraint allows us to mine effective feature representations and boost the segmentation performance. To produce more accurate segmentation maps, we propose a Dual-scale Complementary Fusion (DCF) module that integrates features from two scale-specific decoders operating at different scales. Extensive experimental results on polyp segmentation from colonoscopy images, skin lesion segmentation from dermoscopic images, and Brain glioma segmentation from MRI images demonstrate the effectiveness of our DEC-Seg against other state-of-the-art semi-supervised segmentation approaches. © 2024, The Authors. All rights reserved.

关键词： computer aided diagnosis

来源：评论

学校读者我要写书评

暂无评论

A Survey on Deep Learning for Polyp Segmentation: Techniques, Challenges and Future Trends

arXiv

引用

arXiv 2023年

作者： Mei, Jiaxin Zhou, Tao Huang, Kaiwen Zhang, Yizhe Zhou, Yi Wu, Ye Fu, Huazhu Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing China School of Computer Science and Engineering Southeast University Nanjing China Institute of High Performance Computing A*STAR Singapore

Early detection and assessment of polyps play a crucial role in the prevention and treatment of colorectal cancer (CRC). Polyp segmentation provides an effective solution to assist clinicians in accurately locating and segmenting polyp regions. In the past, people often relied on manually extracted lower-level features such as color, texture, and shape, which often had problems capturing global context and lacked robustness to complex scenarios. With the advent of deep learning, more and more medical image segmentation algorithms based on deep learning networks have emerged, making significant progress in the field. This paper provides a comprehensive review of polyp segmentation algorithms. We first review some traditional algorithms based on manually extracted features and deep segmentation algorithms, and then describe benchmark datasets related to the topic. Specifically, we carry out a comprehensive evaluation of recent deep learning models and results based on polyp size, taking into account the focus of research topics and differences in network structures. Finally, we discuss the challenges of polyp segmentation and future trends in the field. Copyright © 2023, The Authors. All rights reserved.

关键词： Medical imaging

来源：评论

学校读者我要写书评

暂无评论

ViLa-MIL: Dual-scale Vision-Language Multiple Instance Learning for Whole Slide Image Classification

ViLa-MIL: Dual-scale Vision-Language Multiple Instance Learn...

引用

Conference on computer Vision and Pattern Recognition (CVPR)

作者： Jiangbo Shi Chen Li Tieliang Gong Yefeng Zheng Huazhu Fu School of Computer Science and Technology Xi'an Jiaotong University Xi'an China Tencent YouTu Lab Jarvis Research Center Shenzhen China Agency for Science Technology and Research Institute of High Performance Computing Singapore

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

Multiple instance learning (MIL)-based framework has become the mainstream for processing the whole slide image (WSI) with giga-pixel size and hierarchical image context in digital pathology. However, these methods heavily depend on a substantial number of bag-level labels and solely learn from the original slides, which are easily affected by variations in data distribution. Recently, vision language model (VLM)-based methods introduced the language prior by pre-training on large-scale pathological image-text pairs. However, the previous text prompt lacks the consideration of pathological prior knowledge, there-fore does not substantially boost the model's performance. Moreover, the collection of such pairs and the pre-training process are very time-consuming and source-intensive. To solve the above problems, we propose a dual-scale vision-language multiple instance learning (ViLa-MIL) framework for whole slide image classification. Specifically, we propose a dual-scale visual descriptive text prompt based on the frozen large language model (LLM) to boost the performance of VLM effectively. To transfer the VLM to process WSI efficiently, for the image branch, we propose a prototype-guided patch decoder to aggregate the patch features progressively by grouping similar patches into the same prototype; for the text branch, we introduce a context-guided text decoder to enhance the text features by incorporating the multi-granular image contexts. Extensive studies on three multi-cancer and multi-center subtyping datasets demonstrate the superiority of ViLa-MIL.

关键词： Pathology Visualization computer vision Aggregates Large language models Computational modeling Prototypes

来源：评论

学校读者我要写书评

暂无评论

Differentiable Hash Encoding for Physics-Informed Neural Networks

Differentiable Hash Encoding for Physics-Informed Neural Net...

引用

Artificial Intelligence (CAI), IEEE Conference on

作者： Ge Jin Deyou Wang Jian Cheng Wong Shipeng Li School of Aerospace Engineering (SAE) Beijing Institute of Technology (BIT) Beijing China School of Computer Science and Engineering (SCSE) Nanyang Technological University (NTU) Singapore Institute of High Performance Computing (IHPC) Agency for Science Technology and Research (A*STAR) Singapore

ISBN: (数字)9798350354096

ISBN: (纸本)9798350354102

Physics-informed neural networks (PINNs) have received considerable attention in the field of scientific computing. Enhancing their performance to fully realize their potential is a key concern in related fields. Recent studies have shown that multiresolution hash encoding can significantly improve the training performance of neural networks, which has been well-documented in various neural representation tasks. However, the global non-differentiable nature of widely used linear interpolation hash encoding makes it unsuitable for direct combination with automatic differentiation (AD) based PINNs. This work introduces and analyzes two differentiable hash encoding methods and studies their performance through numerical experiments. The proposed encoding methods are combined directly with AD-based PINNs, which, to the best of our knowledge, has not been done before.

关键词： Training Interpolation Scientific computing Neural networks Encoding Task analysis Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

PancakeFS: A Write Efficiently and Read Optimized Filesystem

PancakeFS: A Write Efficiently and Read Optimized Filesystem

引用

Consumer Electronics and computer engineering (ICCECE), 2021 IEEE International Conference on

作者： Yongkang Wang Yang Liu Zhiguang Chen School of Computer Science and Engineering Sun Yat-sen University Guangzhou China State Key Laboratory of High-Performance Computing Institute for Quantum Information College of Computer National University of Defense Technology Changsha China

With the development of emerging technologies such as cloud computing and large AI models (such as LLM), many applications have placed higher demands on the intensive read and write processing of massive data. Addressing the limitations of traditional file systems, we propose PancakeFS, an LSM-Tree-based file system. Leveraging the write-friendly characteristics of LSM-Tree, PancakeFS exhibits efficient write performance. We improve read performance through the design of an inode allocation algorithm based on subtree partitioning and optimization of the disk layout for hot data reconstruction. Our experimental results demonstrate that PancakeFS outperforms Ext4 and TableFS in terms of both read and write performance.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Hybrid Convolutional Neural Network with Change Detection on Hyperspectral Imagery 1

Hybrid Convolutional Neural Network with Change Detection on...

引用

1st International Conference on Industrial Electronics Research and Applications, ICIERA 2021

作者： Bidari, Indira Chickerur, Satyadhyan Thm, Abhishek School of Computer Science and Engineering KLE Technological University Hubballi India Centre for High Performance Computing KLE Technological University Hubballi India

ISBN: (纸本)9781665435420

Change Detection (CD) is one of the major research areas in the field of remote sensing. Hyperspectral Images (HSI's) boosted the change detection technology with their high spectral resolution features. Traditional CD techniques process the low dimensional images not well suited for high dimensional HSI's. Even tiny details can be captured using hyperspectral images, but processing these images is difficult because of their complex high-dimensional data. HSI contains noise and redundancy that affect the spectral features of hyperspectral imagery. We present a hybrid convolution neural network method to address processing complex high dimensional data of hyperspectral images. The proposed network extracts the spectral-spatial information of hyperspectral images by decomposing the change Tensor in three directions. The spectral dimension of the change Tensor is reduced by using I-D Convolution' and then the Tensor is decomposed from two spatial dimensions. A 2-D convolution is applied to extract the spectral and spatial features along different spatial dimensions to improve accuracy. The results on three hyperspectral image datasets illustrate the performance improvement than most state-of-The-Art techniques. © 2021 IEEE.

关键词： Change detection

来源：评论

学校读者我要写书评

暂无评论

Soft Constraint in Local Structure Approximation-PINN

Soft Constraint in Local Structure Approximation-PINN

引用

Artificial Intelligence (CAI), IEEE Conference on

作者： Jian Cheng Wong Pao-Hsiung Chiu Chinchun Ooi My Ha Dao Institute of High Performance Computing Agency for Science Technology and Research (A*STAR) Singapore School of Computer Science and Engineering Nanyang Technological University (NTU) Singapore Center for Frontier AI Research Agency for Science Technology and Research (A*STAR) Singapore

ISBN: (数字)9798350354096

ISBN: (纸本)9798350354102

We previously presented a novel loss formulation for efficient learning of complex dynamics from governing physics, typically described by partial differential equations (PDEs), using physics-informed neural networks (PINNs). In these experiments, the incorporation of a Boundary Connectivity (BCXN) loss function was shown to greatly improve physics-informed learning across many problems, especially those with complex geometries. However, this imposition may not always be ideal, as the previously presented BCXN-loss strongly enforces a linear local structure at the boundary. While this assumption helps facilitate faster learning with an order of magnitude sparser training samples, this can adversely impact convergence in other situations. Hence, we propose a modification of this BCXN-loss that reduces the imposed structure to a soft constraint, allowing for more flexible learning and convergence. We further demonstrate the potential for this method to improve the convergence and performance of LSA-PINN across additional numerical experiments, with much smaller errors than existing methods in terms of the standard L2-norm metric. In particular, we have applied this method to the modelling of flow past complex geometries such as airfoils, which serve as the basic building block for various applications in fluid dynamics and renewable energy (e.g., in wind and tidal turbine design).

关键词： Geometry Training Renewable energy sources Wind energy Fluid dynamics Wind turbines Wind forecasting

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：