This paper first puts forward a case based system framework based on data mining techniques. Then the paper examines the possibility of using neural networks as a method of retrieval in such a case based system. In ...This paper first puts forward a case based system framework based on data mining techniques. Then the paper examines the possibility of using neural networks as a method of retrieval in such a case based system. In this system we propose data mining algorithms to discover case knowledge and other algorithms.展开更多
文本-图像行人检索(text-based person retrieval)作为多模态智能监控系统的核心任务,旨在通过自由形式的文本描述从大规模数据库中识别目标行人图像,在公共安全与视频取证领域具有关键应用价值,如刑事侦查中的嫌疑人追踪及跨摄像头取...文本-图像行人检索(text-based person retrieval)作为多模态智能监控系统的核心任务,旨在通过自由形式的文本描述从大规模数据库中识别目标行人图像,在公共安全与视频取证领域具有关键应用价值,如刑事侦查中的嫌疑人追踪及跨摄像头取证分析.传统方法通常基于图像-文本对完美对齐的理想化假设,忽视了实际场景中普遍存在的复杂噪声数据问题,即视觉实例与其文本标注间因人工标注偏差、网络爬取噪声,或局部视觉属性与全局文本语境间的语义粒度失配而产生的错误或歧义性关联.为弥补这一缺陷,提出了一种语义感知噪声关联学习框架,通过双重创新机制系统性地实现噪声辨识与鲁棒学习.首先,语义感知噪声辨识准则融合模态内语义一致性与跨模态交互信号,基于自适应阈值判定精准区分噪声关联;其次,噪声鲁棒互补学习范式实施差异化优化策略:对于可靠子集采用对比损失进行正向学习以增强特征判别性,而对噪声子集则通过反向学习以抑制过拟合.在3个公开基准数据集上的大量实验表明,该方法在合成噪声数据与真实噪声数据场景中均展现出优越性能.展开更多
This paper describes a building subsidence deformation prediction model with the self-memorization principle.According to the non-linear specificity and monotonic growth characteristics of the time series of building ...This paper describes a building subsidence deformation prediction model with the self-memorization principle.According to the non-linear specificity and monotonic growth characteristics of the time series of building subsidence deformation,a data-based mechanistic self-memory model considering randomness and dynamic features of building subsidence deformation is established based on the dynamic data retrieved method and the self-memorization equation.This model first deduces the differential equation of the building subsidence deformation system using the dynamic retrieved method,which treats the monitored time series data as particular solutions of the nonlinear dynamic system.Then,the differential equation is evolved into a difference-integral equation by the self-memory function to establish the self-memory model of dynamic system for predicting nonlinear building subsidence deformation.As the memory coefficients of the proposed model are calculated with historical data,which contain useful information for the prediction and overcome the shortcomings of the average prediction,the model can predict extreme values of a system and provide higher fitting precision and prediction accuracy than deterministic or random statistical prediction methods.The model was applied to subsidence deformation prediction of a building in Xi'an.It was shown that the model is valid and feasible in predicting building subsidence deformation with good accuracy.展开更多
In recent years there has been a significant interest in peer-to-peer （P2P） environments in the community of data management. However, almost all work, so far, is focused on exact query processing in current P2P dat...In recent years there has been a significant interest in peer-to-peer （P2P） environments in the community of data management. However, almost all work, so far, is focused on exact query processing in current P2P data systems. The autonomy of peers also is not considered enough. In addition, the system cost is very high because the information publishing method of shared data is based on each document instead of document set. In this paper, abstract indices （AbIx） are presented to implement content-based approximate queries in centralized, distributed and structured P2P data systems. It can be used to search as few peers as possible but get as many returns satisfying users＇ queries as possible on the guarantee of high autonomy of peers. Also, abstract indices have low system cost, can improve the query processing speed, and support very frequent updates and the set information publishing method. In order to verify the effectiveness of abstract indices, a simulator of 10,000 peers, over 3 million documents is made, and several metrics are proposed. The experimental results show that abstract indices work well in various P2P data systems.展开更多
基金Supported by the National Science of China(6 0 0 75 0 15 ) and Key Project of Scientific and Technological Departmentin Anhui
文摘This paper first puts forward a case based system framework based on data mining techniques. Then the paper examines the possibility of using neural networks as a method of retrieval in such a case based system. In this system we propose data mining algorithms to discover case knowledge and other algorithms.
文摘文本-图像行人检索(text-based person retrieval)作为多模态智能监控系统的核心任务,旨在通过自由形式的文本描述从大规模数据库中识别目标行人图像,在公共安全与视频取证领域具有关键应用价值,如刑事侦查中的嫌疑人追踪及跨摄像头取证分析.传统方法通常基于图像-文本对完美对齐的理想化假设,忽视了实际场景中普遍存在的复杂噪声数据问题,即视觉实例与其文本标注间因人工标注偏差、网络爬取噪声,或局部视觉属性与全局文本语境间的语义粒度失配而产生的错误或歧义性关联.为弥补这一缺陷,提出了一种语义感知噪声关联学习框架,通过双重创新机制系统性地实现噪声辨识与鲁棒学习.首先,语义感知噪声辨识准则融合模态内语义一致性与跨模态交互信号,基于自适应阈值判定精准区分噪声关联;其次,噪声鲁棒互补学习范式实施差异化优化策略:对于可靠子集采用对比损失进行正向学习以增强特征判别性,而对噪声子集则通过反向学习以抑制过拟合.在3个公开基准数据集上的大量实验表明,该方法在合成噪声数据与真实噪声数据场景中均展现出优越性能.
基金supported by the Twelfth Five National Key Technology R&D Program of China (2009BAJ28B04,2011BAK07B01,2011BAJ08B03,2011BAJ08B05)the National Natural Science Foundation of China(51108428)+1 种基金Beijing Postdoctoral Research Foundation (2012ZZ-17)China Postdoctoral Science Foundation (2011M500199)
文摘This paper describes a building subsidence deformation prediction model with the self-memorization principle.According to the non-linear specificity and monotonic growth characteristics of the time series of building subsidence deformation,a data-based mechanistic self-memory model considering randomness and dynamic features of building subsidence deformation is established based on the dynamic data retrieved method and the self-memorization equation.This model first deduces the differential equation of the building subsidence deformation system using the dynamic retrieved method,which treats the monitored time series data as particular solutions of the nonlinear dynamic system.Then,the differential equation is evolved into a difference-integral equation by the self-memory function to establish the self-memory model of dynamic system for predicting nonlinear building subsidence deformation.As the memory coefficients of the proposed model are calculated with historical data,which contain useful information for the prediction and overcome the shortcomings of the average prediction,the model can predict extreme values of a system and provide higher fitting precision and prediction accuracy than deterministic or random statistical prediction methods.The model was applied to subsidence deformation prediction of a building in Xi'an.It was shown that the model is valid and feasible in predicting building subsidence deformation with good accuracy.
基金Supported by the National Natural Science Foundation of China under Grant No. 60473077 and the Program for New Century Excellent Talents in University.
文摘In recent years there has been a significant interest in peer-to-peer （P2P） environments in the community of data management. However, almost all work, so far, is focused on exact query processing in current P2P data systems. The autonomy of peers also is not considered enough. In addition, the system cost is very high because the information publishing method of shared data is based on each document instead of document set. In this paper, abstract indices （AbIx） are presented to implement content-based approximate queries in centralized, distributed and structured P2P data systems. It can be used to search as few peers as possible but get as many returns satisfying users＇ queries as possible on the guarantee of high autonomy of peers. Also, abstract indices have low system cost, can improve the query processing speed, and support very frequent updates and the set information publishing method. In order to verify the effectiveness of abstract indices, a simulator of 10,000 peers, over 3 million documents is made, and several metrics are proposed. The experimental results show that abstract indices work well in various P2P data systems.
