Survey of Clustering Methods and Related Seeding Techiniques
نویسنده
چکیده
We try to extend the seeding technique proposed in [2],to the spectral clustering domain.We start with the survey of a sequence of papers on the connection between spectral clustering techniques and centroid-based algorithms. We show that the results in [2] can provide a bound for the rounding of the normalized cut problem. This raises an interesting question on the relationship between the seeding techinique proposed by Arthur, and the spectral relaxation of the normalized cut problem.
منابع مشابه
Fujitsu Laboratories Trec7 Report 2 System Description 2.1 Overall 2.2 the Search System Tera
In our rst participation in TREC, our focus was on improving the basic ranking systems and applying text clustering techniques for query expansion. We tested a variety of techiniques including reference measures, passage retrieval, and data fusion for the basic ranking systems. Some techiniques were used in the o cial run, others were not used because of time limitations. We applied the text cl...
متن کاملOPTIMIZATION OF FUZZY CLUSTERING CRITERIA BY A HYBRID PSO AND FUZZY C-MEANS CLUSTERING ALGORITHM
This paper presents an efficient hybrid method, namely fuzzy particleswarm optimization (FPSO) and fuzzy c-means (FCM) algorithms, to solve the fuzzyclustering problem, especially for large sizes. When the problem becomes large, theFCM algorithm may result in uneven distribution of data, making it difficult to findan optimal solution in reasonable amount of time. The PSO algorithm does find ago...
متن کاملCentralized Clustering Method To Increase Accuracy In Ontology Matching Systems
Ontology is the main infrastructure of the Semantic Web which provides facilities for integration, searching and sharing of information on the web. Development of ontologies as the basis of semantic web and their heterogeneities have led to the existence of ontology matching. By emerging large-scale ontologies in real domain, the ontology matching systems faced with some problem like memory con...
متن کاملImproved Automatic Clustering Using a Multi-Objective Evolutionary Algorithm With New Validity measure and application to Credit Scoring
In data mining, clustering is one of the important issues for separation and classification with groups like unsupervised data. In this paper, an attempt has been made to improve and optimize the application of clustering heuristic methods such as Genetic, PSO algorithm, Artificial bee colony algorithm, Harmony Search algorithm and Differential Evolution on the unlabeled data of an Iranian bank...
متن کاملEffects of different tillage system, seeding method and rates on yield and seed oil percentage of rapeseed
Rapeseed as one the most crops is the basic needs of human feeding in order to produce oil. A field experiment was conducted to study the effects of different tillage system, seeding method and rates on yield and its components, as well as, seed oil percentage of rapeseed (Brassica napus L. cv. Hyola-401) during 2011-12 growing season at Rice Research Institute of Iran-Department of Ma...
متن کامل