检索结果-内蒙古大学图书馆

Modeling topical relevance for multi-turn dialogue generation 29

学校读者我要写书评

暂无评论

Modeling topical relevance for multi-turn dialogue generatio...

29th International Joint Conference on Artificial Intelligence, IJCAI 2020

作者： Zhang, Hainan Lan, Yanyan Pang, Liang Chen, Hongshen Ding, Zhuoye Yin, Dawei CAS Key Lab of Network Data Science and Technology Institute of Computing Technology CAS China *** Beijing China University of Chinese Academy of Sciences China

ISBN: (纸本)9780999241165

Topic drift is a common phenomenon in multi-turn dialogue. Therefore, an ideal dialogue generation models should be able to capture the topic information of each context, detect the relevant context, and produce appropriate responses accordingly. However, existing models usually use word or sentence level similarities to detect the relevant contexts, which fail to well capture the topical level relevance. In this paper, we propose a new model, named STAR-BTM, to tackle this problem. Firstly, the Biterm Topic Model is pre-trained on the whole training dataset. Then, the topic level attention weights are computed based on the topic representation of each context. Finally, the attention weights and the topic distribution are utilized in the decoding process to generate the corresponding responses. Experimental results on both Chinese customer services data and English Ubuntu dialogue data show that STAR-BTM significantly outperforms several state-of-the-art methods, in terms of both metric-based and human evaluations. © 2020 Inst. Sci. inf., Univ. Defence in Belgrade. All rights reserved.

关键词： Stars

Understanding and Improving Neural Ranking Models from a Term Dependence View 15th

学校读者我要写书评

暂无评论

Understanding and Improving Neural Ranking Models from a Ter...

15th Asia Information Retrieval Societies Conference, AIRS 2019

作者： Fan, Yixing Guo, Jiafeng Lan, Yanyan Cheng, Xueqi CAS Key Lab of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China

ISBN: (纸本)9783030428341

Recently, neural information retrieval (NeuIR) has attracted a lot of interests, where a variety of neural models have been proposed for the core ranking problem. Beyond the continuous refresh of the state-of-the-art neural ranking performance, the community calls for more analysis and understanding of the emerging neural ranking models. In this paper, we attempt to analyze these new models from a traditional view, namely term dependence. Without loss of generality, most existing neural ranking models could be categorized into three categories with respect to their underlying assumption on query term dependence, i.e., independent models, dependent models, and hybrid models. We conduct rigorous empirical experiments over several representative models from these three categories on a benchmark dataset and a large click-through dataset. Interestingly, we find that no single type of model can achieve a consistent win over others on different search queries. An oracle model which can select the right model for each query can obtain significant performance improvement. Based on the analysis we introduce an adaptive strategy for neural ranking models. We hypothesize that the term dependence in a query could be measured through the divergence between its independent and dependent representations. We thus propose a dependence gate based on such divergence representation to softly select neural ranking models for each query accordingly. Experimental results verify the effectiveness of the adaptive strategy. © Springer Nature Switzerland AG 2020.

关键词： Large dataset

Hard Negatives or False Negatives: Correcting Pooling Bias in Training Neural Ranking Models

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Cai, Yinqiong Guo, Jiafeng Fan, Yixing Ai, Qingyao Zhang, Ruqing Cheng, Xueqi CAS Key Lab of Network Data Science and Technology ICT CAS University of Chinese Academy of Sciences Beijing China Dept. CS&T Beijing National Research Center for Information Science and Technology Tsinghua University Beijing China

Neural ranking models (NRMs) have become one of the most important techniques in information retrieval (IR). Due to the limitation of relevance labels, the training of NRMs heavily relies on negative sampling over unlabeled data. In general machine learning scenarios, it has shown that training with hard negatives (i.e., samples that are close to positives) could lead to better performance. Surprisingly, we find opposite results from our empirical studies in IR. When sampling top-ranked results (excluding the labeled positives) as negatives from a stronger retriever, the performance of the learned NRM becomes even worse. Based on our investigation, the superficial reason is that there are more false negatives (i.e., unlabeled positives) in the top-ranked results with a stronger retriever, which may hurt the training process;The root is the existence of pooling bias in the dataset constructing process, where annotators only judge and label very few samples selected by some basic retrievers. Therefore, in principle, we can formulate the false negative issue in training NRMs as learning from labeled datasets with pooling bias. To solve this problem, we propose a novel Coupled Estimation Technique (CET) that learns both a relevance model and a selection model simultaneously to correct the pooling bias for training NRMs. Empirical results on three retrieval benchmarks show that NRMs trained with our technique can achieve significant gains on ranking effectiveness against other baseline strategies. © 2022, CC0.

关键词：

Augmentation-aware self-supervision for data-efficient GAN training 23

学校读者我要写书评

暂无评论

Augmentation-aware self-supervision for data-efficient GAN t...

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

作者： Liang Hou Qi Cao Yige Yuan Songtao Zhao Chongyang Ma Siyuan Pan Pengfei Wan Zhongyuan Wang Huawei Shen Xueqi Cheng CAS Key Laboratory of AI Safety and Security Institute of Computing Technology Chinese Academy of Sciences and University of Chinese Academy of Sciences and Kuaishou Technology CAS Key Laboratory of AI Safety and Security Institute of Computing Technology Chinese Academy of Sciences CAS Key Laboratory of AI Safety and Security Institute of Computing Technology Chinese Academy of Sciences and University of Chinese Academy of Sciences Kuaishou Technology CAS Key Laboratory of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences and University of Chinese Academy of Sciences

Training generative adversarial networks (GANs) with limited data is challenging because the discriminator is prone to overfitting. Previously proposed differentiable augmentation demonstrates improved data efficiency of training GANs. However, the augmentation implicitly introduces undesired invariance to augmentation for the discriminator since it ignores the change of semantics in the label space caused by data transformation, which may limit the representation learning ability of the discriminator and ultimately affect the generative modeling performance of the generator. To mitigate the negative impact of invariance while inheriting the benefits of data augmentation, we propose a novel augmentation-aware self-supervised discriminator that predicts the augmentation parameter of the augmented data. Particularly, the prediction targets of real data and generated data are required to be distinguished since they are different during training. We further encourage the generator to adversarially learn from the self-supervised discriminator by generating augmentation-predictable real and not fake data. This formulation connects the learning objective of the generator and the arithmetic - harmonic mean divergence under certain assumptions. We compare our method with state-of-the-art (SOTA) methods using the class-conditional BigGAN and unconditional StyleGAN2 architectures on data-limited CIFAR-10, CIFAR-100, FFHQ, LSUN-Cat, and five low-shot datasets. Experimental results demonstrate significant improvements of our method over SOTA methods in training data-efficient GANs. Our code is available at https://***/liang-hou/augself-gan.

关键词：

DAPO-QAOA: An algorithm for solving combinatorial optimization problems by dynamically constructing phase operators

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Wang, Yukun Li, ZeYang Wan, Linchun Beijing Key Laboratory of Petroleum Data Mining China University of Petroleum Beijing102249 China State Key Lab of Processors Institute of Computing Technology CAS Beijing100190 China School of Computer and Information Science Southwest University Chongqing400715 China

The Quantum Approximate Optimization Algorithm (QAOA) is a well-known hybrid quantum-classical algorithm for combinatorial optimization problems. Improving QAOA involves enhancing its approximation ratio while addressing practical constraints of Noisy Intermediate Scale Quantum (NISQ) devices, such as minimizing the number of two-qubit gates and reducing circuit depth. Although existing research has optimized designs for phase and mixer operators to improve performance, challenges remain, particularly concerning the excessive use of two-qubit gates in the construction of phase operators. To address these issues, we introduce a Dynamic Adaptive Phase Operator (DAPO) algorithm, which dynamically constructs phase operators based on the output of previous layers and neighborhood search approach, optimizing the problem Hamiltonian more efficiently. By using solutions generated by QAOA itself to simplify the problem Hamiltonian at each layer, the algorithm captures the problem’s structural properties more effectively, progressively steering the solution closer to the optimal target. Experimental results on MaxCut and NAE3SAT problems show that DAPO achieves higher approximation ratios and significantly reduces two-qubit RZZ gates, especially in dense graphs. Compared to vanilla QAOA, DAPO uses only 66% of RZZ gates at the same depth while delivering better results, demonstrating its potential for efficient combinatorial optimization in the NISQ era. © 2025, CC BY.

关键词： Hamiltonians

An Ensemble of CapsNet and KSVM Using Feature Fusion: Application to COVID-19 Detection 3

学校读者我要写书评

暂无评论

An Ensemble of CapsNet and KSVM Using Feature Fusion: Applic...

3rd International Conference on Pattern Recognition and Machine Learning, PRML 2022

作者： Aggrey, Esther Stacy E. B. Zhen, Qin Kodjiku, Seth Larweh Aidoo, Evans Fiasam, Linda Delali Mensah, Acheampong Edward University of Electronic Science & Technology of China School of Information & Software Engineering Chengdu China University of Electronic Science & Technology of China School of Information & Software Engineering Network and Data Security Key Lab of Sichuan Province Chengdu China Zhejiang Gongshang University School of Computer & Information Engineering Hangzhou China

ISBN: (数字)9781665499507

ISBN: (纸本)9781665499507

COVID-19 virus is a major worldwide pandemic that is growing at a fast pace throughout the world. The usual approach for diagnosing COVID-19 is the use of a real-time polymerase chain reaction (RT-PCR) based nucleic acid test. However, RT-PCR has lower sensitivity in the early phases of COVID-19 detection. Recent studies have indicated that X-ray images may be useful throughout the early detection of the virus. Human screening has been shown to be cost-effective, susceptible to mistakes, and time-demanding, which has sparked an interest in using Convolutional Neural networks (CNNs) to automate the process. CNNs, on the other hand, fail to view the exact placement of features as advantageous in medical imaging. Furthermore, for successful training and prediction, CNNs need a huge quantity of datasets. CNNs are rapidly reducing picture resolution, resulting in worsening accuracy in classification. We used newly created capsule networks (CapsNets) in our study to circumvent these disadvantages. The primary contribution is to improve the identification of SARS-CoV-2 with images obtained from X-ray by coupling capsule network with a kernel support vector machine (KSVM). The technique was evaluated using a publicly available dataset, and the proposed model shows that the accuracy of the CapsNet-KSVM based model is improved by 94.6% accuracy, 95% sensitivity, and 98% specificity, which outperforms the traditional CNN and other existing ensemble models. The proposed CapsNet-KSVM based system can be employed to identify the presence of COVID-19 in the human body using X-ray images. © 2022 IEEE.

关键词： COVID-19

FRCNet: Frequency and Region Consistency for Semi-supervised Medical Image Segmentation

学校读者我要写书评

暂无评论

arXiv 2024年

作者： He, Along Li, Tao Wu, Yanlin Zou, Ke Fu, Huazhu Tianjin Key Laboratory of Network and Data Security Technology College of Computer Science Nankai University Tianjin China Haihe Lab of ITAI China National Key Laboratory of Fundamental Science on Synthetic Vision Sichuan University Sichuan Chengdu China Institute of High Performance Computing Agency for Science Technology and Research Singapore

Limited labeled data hinder the application of deep learning in medical domain. In clinical practice, there are sufficient unlabeled data that are not effectively used, and semi-supervised learning (SSL) is a promising way for leveraging these unlabeled data. However, existing SSL methods ignore frequency domain and region-level information and it is important for lesion regions located at low frequencies and with significant scale changes. In this paper, we introduce two consistency regularization strategies for semi-supervised medical image segmentation, including frequency domain consistency (FDC) to assist the feature learning in frequency domain and multi-granularity region similarity consistency (MRSC) to perform multi-scale region-level local context information feature learning. With the help of the proposed FDC and MRSC, we can leverage the powerful feature representation capability of them in an effective and efficient way. We perform comprehensive experiments on two datasets, and the results show that our method achieves large performance gains and exceeds other state-of-the-art methods. © 2024, CC BY.

关键词： Deep learning

LoL: A Comparative Regularization Loss over Query Reformulation Losses for Pseudo-Relevance Feedback

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Zhu, Yunchang Pang, Liang Lan, Yanyan Shen, Huawei Cheng, Xueqi Data Intelligence System Research Center Institute of Computing Technology Cas University of Chinese Academy of Sciences Beijing China Data Intelligence System Research Center Institute of Computing Technology Cas Beijing China Institute for Ai Industry Research Tsinghua University Beijing China Cas Key Lab of Network Data Science and Technology Institute of Computing Technology Cas University of Chinese Academy of Sciences Beijing China

Pseudo-relevance feedback (PRF) has proven to be an effective query reformulation technique to improve retrieval accuracy. It aims to alleviate the mismatch of linguistic expressions between a query and its potential relevant documents. Existing PRF methods independently treat revised queries originating from the same query but using different numbers of feedback documents, resulting in severe query drift. Without comparing the effects of two different revisions from the same query, a PRF model may incorrectly focus on the additional irrelevant information increased in the more feedback, and thus reformulate a query that is less effective than the revision using the less feedback. Ideally, if a PRF model can distinguish between irrelevant and relevant information in the feedback, the more feedback documents there are, the better the revised query will be. To bridge this gap, we propose the Loss-over-Loss (LoL) framework to compare the reformulation losses between different revisions of the same query during training. Concretely, we revise an original query multiple times in parallel using different amounts of feedback and compute their reformulation losses. Then, we introduce an additional regularization loss on these reformulation losses to penalize revisions that use more feedback but gain larger losses. With such comparative regularization, the PRF model is expected to learn to suppress the extra increased irrelevant information by comparing the effects of different revised queries. Further, we present a differentiable query reformulation method to implement this framework. This method revises queries in the vector space and directly optimizes the retrieval performance of query vectors, applicable for both sparse and dense retrieval models. Empirical evaluation demonstrates the effectiveness and robustness of our method for two typical sparse and dense retrieval models. © 2022, CC BY.

关键词： Feedback

Continuous Distributed Processing of Software Defined Radar

学校读者我要写书评

暂无评论

Continuous Distributed Processing of Software Defined Radar

2021 CIE International Conference on Radar, Radar 2021

作者： Li, Bing Qiu, Qiang Gong, Shiqi Liu, Yongjun Lei, Yu Institute of Computing Technology CAS Key Lab of Network Data Science and Technology Chinese Academy of Sciences Beijing China State Key Laboratory of Internet of Things for Smart City University of Macau Macau China Xidian University National Laboratory of Radar Signal Processing Xi'an China Golaxy Data Technology Co. Ltd. Beijing China

ISBN: (纸本)9781665498142

Software-defined radar has been an active research field for more than ten years. However, the low performance and low scalability of the traditional processing techniques of SDR make it hard to deal with complex radar tasks. The stream processing engine is a software that supports low latency and high scalability distributed data processing. It is widely used in log, image, and video processing scenarios but rarely used in radar applications. This paper reviews processing software-defined radar data by using the stream processing engine, and we illustrate the differences between processing log data and radar data. Furthermore, we discuss some optimizations for stream processing engines to get better performance in the radar data processing. © 2021 IEEE.

关键词： distributed processing Software-defined radar stream processing engine