A Modified Fuzzy K-means Clustering using Expectation Maximization

Nasser, Sara; Alkhaldi, R.; Vert, Gregory

doi:10.1109/fuzzy.2006.1681719

Cited by 41 publications

(23 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…G. Chicco and his colleagues have presented several papers such as [7,[15][16][17][18] which discuss different clustering techniques for load curve classification including K-means, hierarchical algorithms, modified follow the leader, etc. A fusion of fuzzy K-means and the expectation maximization algorithm method is discussed in [19]. Also, [20] improves K-means by determining new greedy initialization method.…”

Section: Literature Reviewmentioning

confidence: 99%

An annual framework for clustering-based pricing for an electricity retailer

Mahmoudi-Kohan¹,

Moghaddam²,

Sheikh‐El‐Eslami³

2010

Electric Power Systems Research

View full text Add to dashboard Cite

Section: Literature Reviewmentioning

confidence: 99%

An annual framework for clustering-based pricing for an electricity retailer

Mahmoudi-Kohan¹,

Moghaddam²,

Sheikh‐El‐Eslami³

2010

Electric Power Systems Research

View full text Add to dashboard Cite

“…For comparison purposes, several K-means based methods available in the literature have also been applied to the dataset, including classical K-means [4], [5], [9], [11]- [18], [20], [21], fuzzy K-means [11]- [14], [17], [18], [27]- [29], K-means++ [30], [31], weighted fuzzy average K-means (WFA-K-means) [16], [29], [32], and developed K-means [17]. Moreover, the H-K-means methods with the other numbers of levels (L = 3, 4, 5, 7) are also included.…”

Section: Performance Comparisonsmentioning

confidence: 99%

Hierarchical K-means Method for Clustering Large-Scale Advanced Metering Infrastructure Data

Chiang

Liu

et al. 2017

IEEE Trans. Power Delivery

View full text Add to dashboard Cite

Clustering of the load patterns from distribution network customers is of vital importance for several applications. However, the growing number of Advanced Metering Infrastructure (AMI) and a variety of customer behaviors make the clustering task quite challenging due to the increasing amount of load data. K-means is a widely used clustering algorithm in processing large dataset with acceptable computational efficiency, but suffers from local optimal solutions. To address this issue, this paper presents a hierarchical K-means (H-K-means) method for better clustering performance for big data problems. The proposed method is applied to a large-scale AMI dataset and its effectiveness is evaluated by benchmarking with several existing clustering methods in terms of five common adequacy indices, outliers detection, and computation time.Index Terms-Advanced Metering Infrastructure (AMI), big data problems, clustering, hierarchical K-means (H-K-means), load patterns.

show abstract

“…Also, for clustering, the optimal number of clusters to be chosen remains a problem. In addition, clustering may fail to converge in finite steps (Nasser et al 2006). Furthermore these algorithms are not scalable.…”

Section: Introductionmentioning

confidence: 95%

Satellite image retrieval using low memory locality sensitive hashing in Euclidean space

Buaba¹,

Homaifar²,

Gebril³

et al. 2011

Earth Sci Inform

View full text Add to dashboard Cite

This paper presents the use of the Low Memory Locality Sensitive Hashing (LMLSH) technique operating in Euclidean space to build a data structure for the Defense Meteorological Satellite Program (DMSP) satellite imagery database. The LMLSH technique finds satellite image matches in sublinear search time. The texture feature vectors of the images are extracted using pyramid-structured wavelet transform coupled with Gaussian central moment technique. These feature vectors and families of hash functions, drawn randomly and independently from a Gaussian distribution, are used to build hash tables. Given a query, the hash tables are used to pull out the best matches to that query and this is done in a sublinear search time complexity. When tested, our algorithm has proven to be approximately twenty six times faster than the Linear Search (LS) algorithm. In addition, the LMLSH algorithm searches about two percent of the entire database randomly to find the possible matches to any given query without loss of accuracy compared to the absolute best matches returned by its LS counterpart.

show abstract

A Modified Fuzzy K-means Clustering using Expectation Maximization

Cited by 41 publications

References 15 publications

An annual framework for clustering-based pricing for an electricity retailer

An annual framework for clustering-based pricing for an electricity retailer

Hierarchical K-means Method for Clustering Large-Scale Advanced Metering Infrastructure Data

Satellite image retrieval using low memory locality sensitive hashing in Euclidean space

Contact Info

Product

Resources

About