18,840 research outputs found
Constrained Data Clustering
No full text近年來，由於各種應用中快速累積了大量資料，資料探勘相關的研究領域越來越受到重視，而其中的資料叢集分析技術，則提供了使用者觀察相似資料群集的途徑。
由於資料探勘的研究常因應用領域而異，其中限制性探勘技術是將應用領域之專業知識加入資料探勘分析考量中的一種方式，在此論文之中的第一個研究課題，即為提出新的限制性資料叢集定義：同一個叢集之中的任意兩個成員，其限制性屬性的差值不可超過所給定的限制範圍。根據此定義，我們提出了幾個相對應的限制性資料叢集演算法，接著，由於觀察到階層式叢集演算法具有的一個基本特性，即資料分群的順序會影響最後的叢集成果，因此又更進一步地設計了漸進式解除限制（progressive constraint relaxation）之技術，以降低分群順序的影響，並提昇分群的成果。
除了針對靜態資料進行資料叢集演算法的研究之外，我們也探討了資料串流環境中的資料叢集技術。在資料串流環境中，資料通常是快速累積，因此需要利用有限的時間與空間資源，提出有效的解決方案。此論文中，我們提出了一個針對多條資料串流進行叢集分析的架構，此架構包含了兩個階段，第一個階段處理並儲存資料串流，第二個階段則提供動態回應使用者叢集分析需求的機制。
最後，我們將限制性資料叢集技術延伸至資料串流環境中，配合本論文中所提出的限制性資料叢集定義，設計相對應的資料串流儲存架構，以產生符合使用者需求與限制的資料叢集。Among various data mining capabilities, data clustering is a useful technique for group behavior investigation, and is helpful for many applications. Since data mining is an application dependent technology, the information involving domain knowledge is usually imposed on the mining systems as various constraints. In this dissertation, we address the problem of constrained clustering with numerical constraints, in which the constraint attribute values of any two data items in the same cluster are required to be within the corresponding constraint range. Several algorithms are proposed to solve such a clustering problem. It is noted that due to the intrinsic nature of the numerical constrained clustering, there is an order dependency on the process of attaining the clustering, which in many cases degrades the clustering results. In view of this, we devise a progressive constraint relaxation technique to remedy this drawback and improve the overall performance of clustering results.
In addition to clustering on static data sets, the problem of clustering multiple data streams is also addressed in this dissertation. We devise a Clustering on Demand framework, abbreviated as COD framework, to dynamically cluster multiple data streams. The COD framework consists of two phases, i.e., the online maintaining phase and the offline clustering phase. The online maintaining phase provides an efficient mechanism to maintain the summary hierarchies of the data streams with multiple resolutions. On the other hand, an adaptive clustering algorithm is devised for the offline phase to retrieve the approximations of the desired sub-streams from the summary hierarchies according to the clustering queries.
Finally, the concepts of constraints and data streams are combined and considered together. We devise a framework of Constrained Clustering for the Evolving Data Stream, abbreviated as CCDS framework, to cluster the data stream under the pairwise range constraint. Two phases are designed to maintain the data points and to generate clusters respectively.1 Introduction 7
1.1 Motivation and Overview of the Dissertation . . . . . . . . . . . . . . . . . . 7
1.2 Organization of theDissertation . . . . . . . . . . . . . . . . . . . . . . . . . 10
2 Clustering with Pairwise Numerical Constraints 11
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2 Problem Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2.1 Pair-Wise Constrained Clustering . . . . . . . . . . . . . . . . . . . . 16
2.2.2 Related Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3 Algorithms for Pair-Wise Constrained Clustering . . . . . . . . . . . . . . . 19
2.3.1 Partition-Based Clustering . . . . . . . . . . . . . . . . . . . . . . . . 19
2.3.2 Hierarchical Clustering Algorithms . . . . . . . . . . . . . . . . . . . 25
2.4 Progressive Constraint Relaxation . . . . . . . . . . . . . . . . . . . . . . . . 32
2.4.1 OrderDependency . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.4.2 Progressive Constraint Relaxation . . . . . . . . . . . . . . . . . . . . 35
2.5 Extension toMultiple Constraint Attributes . . . . . . . . . . . . . . . . . . 38
2.6 Performance Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.6.1 On Progressive Constraint Relaxation . . . . . . . . . . . . . . . . . . 41
2.6.2 On Partition-Based Clustering . . . . . . . . . . . . . . . . . . . . . . 42
2.6.3 OnHierarchical Clustering . . . . . . . . . . . . . . . . . . . . . . . . 43
2.6.4 Overall Comparison between theseAlgorithms . . . . . . . . . . . . . 45
2.6.5 On the Scalability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3 Adaptive Clustering for Multiple Evolving Streams 51
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.2 Clustering onDemand forMultipleData Streams . . . . . . . . . . . . . . . 54
3.2.1 Framework Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.2.2 Advantages of the CODFramework . . . . . . . . . . . . . . . . . . . 57
3.3 The Framework of COD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.3.1 OnlineMaintaining Phase . . . . . . . . . . . . . . . . . . . . . . . . 58
3.3.2 Offline Clustering Phase . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.3.3 ComplexityAnalyses . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.3.4 Adaptive Use of COD . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.4 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.4.1 Simulation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.4.2 SensitivityAnalyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.4.3 AdaptivityAnalyses . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
3.4.4 CODon Real Dataset . . . . . . . . . . . . . . . . . . . . . . . . . . 82
3.4.5 On the Scalability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4 Constrained Clustering for the Evolving Data Stream 87
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.2 Constrained Clustering for the EvolvingData Stream . . . . . . . . . . . . . 89
4.3 CCDS Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4.3.1 Statistics Reserving Phase . . . . . . . . . . . . . . . . . . . . . . . . 92
4.3.2 Clustering Responding Phase . . . . . . . . . . . . . . . . . . . . . . 101
4.4 Performance Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
4.4.1 On Parameters of Framework CCDS . . . . . . . . . . . . . . . . . . 104
4.4.2 On Clustering Requests . . . . . . . . . . . . . . . . . . . . . . . . . 106
4.4.3 On Scalability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5 Conclusions 11
Adaptive Clustering for Multiple Evolving Streams
No full textIn the data stream environment, the patterns generated at different time instances are different due to data evolution. As
time progresses, the behavior and members of clusters usually change. Hence, clustering continuous data streams allows us to
observe the changes of group behavior. In order to support flexible clustering requirements, we devise in this paper a Clustering on
Demand framework, abbreviated as COD framework, to dynamically cluster multiple data streams. While providing a general
framework of clustering on multiple data streams, the COD framework has two advantageous features, namely, one data scan for
online statistics collection and compact multiresolution approximations, which are designed to address, respectively, the time and the
space constraints in a data stream environment. The COD framework consists of two phases, i.e., the online maintenance phase and
the offline clustering phase. The online maintenance phase provides an efficient mechanism to maintain summary hierarchies of data
streams with multiple resolutions in time linear in both the number of streams and the number of data points in each stream. On the
other hand, an adaptive clustering algorithm is devised for the offline phase to retrieve approximations of desired substreams from
summary hierarchies according to clustering queries. We propose two summarization techniques, based on wavelet and regression
analyses, to construct the summary hierarchies. The regression-based summary hierarchy approximates the data stream more
precisely and provides better clustering results, at the cost of slightly longer time than and twice the storage space as the waveletbased
one. An adaptive version of COD framework is designed to make a selection between a wavelet-based model and a regressionbased
model for building the summary hierarchy. By the adaptive COD, we can obtain clustering results with almost the same quality
as the regression-based COD while using much less storage space for the summary hierarchy. As shown in the complexity analyses
and also validated by our empirical studies, the COD framework performs very efficiently in the data stream environment while
producing clustering results of very high quality
Ru jia si xiang yu xian dai shi jie
No full textBen shu suo shou ru de lun wen shi zhong yang yan jiu yuan zhong guo wen zhe yan jiu suo tui dong de " dang dai ru xue zhu ti yan jiu ji hua " di yi qi zhi bu fen cheng guo. bao gua " fo xue, xi xue yu dang dai xin ru jia -- hong guan de zhe xue kao cha " deng lun we
Images of the Dai : the aesthetics of gender and identity in Xishuangbanna
No full textThis thesis is based on fieldwork carried out m Xishuangbanna Dai Autonomous Prefecture, Yunnan Province, China. The main focus of the work is the Dai people, one of China's fifty-five so called 'Minority Nationalities'. I aim to paint a picture of the complex processes through which Dai ways of being and images of them are created and recreated. This is not to suggest that the Dai constitute a bounded group. Although Chinese official discourse presents a static, rigid picture of the so-called 'Minority Nationalities', I hope to have demonstrated that the everyday experiences of those in Banna are governed by a fluid and dynamic relationality. Images of 'Minority Nationalities' abound in China, these images are multiple and often contradictory. The Dai are known throughout China for their beauty, a beauty often portrayed as highly erotic. In this thesis I explore the implications of this image and the role of the Dai in its formation and continuity. With this in mind I examine the ways that the striking Dai aesthetic is used in the intricate power plays of Xishuangbanna. This work examines aspects of the Dai lived aesthetic and as such it has chapters on tattoo, architecture and feminine beauty. Dai aesthetic knowledge is interlaced with strands of moral, philosophical and cosmological insight, thus this work also includes a chapter on morality, autonomy and cooperation. The penultimate chapter uses vivid ethnography of the Water Splashing festival as a example of play of identities in Xishuangbanna. The Conclusion reiterates that the processes by which images, identities and aesthetic understandings are generated, and by which limits are explored and transgressed in Xishuangbanna are dialogic in character
白金微粒/Ru錯合物/高分子薄磨修飾電極之製備及其電催化應用
No full text[[abstract]]Membrane-metal modified cells are usually applied to the field of membrane cells, such as fuel cells, light cells… etc. Some hydrogen evolution catalysts, for example Pt metal, are used as the cathodes in these cells. Therefore, in this research, we try to dose a series of Ru complexes into the intervals between the cathode and the thin film electrode to improve the hydrogen evolution efficiency of the membrane-metal modified cells.
Ru complex (chosen from Ru(bpy)2phenNH2, Ru(dmb)2 phenNH2, Ru(tmb)2phenNH2, Ru(bpy)2Cl2, Ru(dmb)2Cl2, Ru(tmb)2Cl2) doped Nafion solution was drop-coated onto glassy carbon (GC) electrode and formed a thin film after drying. Then, the GC electrode was immersed into H2PtCl6 solution, and the Pt/Ru complex/polymer modified electrode was obtained by reducing Pt with a DPTB method. Different amount of Pt was electroplated on the GC electrode even at the same conditions (the same potential, the same time interval) when different consistency or kinds of Ru complexes were used, suggesting the influence of the ligands on the red-ox property of Ru complexes. The effective surface area (estimated by CV method) of Pt in the modified electrode is also varied with different species of Ru complexes, which in turn affect on the efficiency of hydrogen evolution.
Furthermore, from the information obtained by SEM and EDS, the alignment and the density of Pt particles growing on the GC electrode are figured out. From the fluorescence lifetime and luminescence spectra, a good electron-transfer is considered to have occurred between D series Ru complexes and Pt modified electrode that explained why a high hydrogen evolution efficiency has been obtained. The modified electrodes are still stable one month after fabricated and their hydrogen evolution efficiency was as good as a newly prepared one.
Tailoring cobalt‐free La <sub>0.5</sub> Sr <sub>0.5</sub> FeO <sub>3‐δ</sub> cathode with a nonmetal cation‐doping strategy for high‐performance proton‐conducting solid oxide fuel cells
Full text linkA nonmetal doping strategy was exploited for the conventional La0.5Sr0.5FeO3-δ (LSF) cathode, allowing high performance for proton-conducting solid oxide fuel cells (H-SOFCs). Unlike previous studies focusing on the utilization of metal oxides as dopants, phosphorus, which is a nonmetal element, was used as the cation dopant for LSF by partially replacing Fe ions to form the new La0.5Sr0.5Fe0.9P0.1O3-δ (LSFP) compound. The H-SOFC using the LSFP cathode showed a two-fold peak power density as compared to that using the LSF cathode. Both experimental studies and first-principle calculations were used to unveil the mechanisms for the high performance of the LSFP cells
- …
