A fast minimum spanning tree algorithm based on K-means

Zhong, Chongquan; Malinen, Mikko; Miao, Duoqian; Fränti, Pasi

doi:10.1016/j.ins.2014.10.012

Cited by 93 publications

(47 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The connection between graph vertices satisfies the minimal sum of the defined edge weights, and the partition of a graph is obtained by removing edges to form various sub-graphs. Some of representative algorithms and recent variants are: exact MST [37], sparse graph [38], dual-tree [39], distributed MST [40], approximate MST [41][42][43], and recursive MST [44]. b) Minimal and normalized graph cut: Graph cut is the partitioning of graph vertices into two disjoint subsets that are joined by at least one edge, and the minimum cut of a graph is the cut set which has the smallest number of unweighted edges or smallest sum of weights possible.…”

Section: Image Segmentationmentioning

confidence: 99%

Semantic content-based image retrieval: A comprehensive study

Alzu’bi

Amira

Ramzan

2015

Journal of Visual Communication and Image Representation

180

106

View full text Add to dashboard Cite

Section: Image Segmentationmentioning

confidence: 99%

Semantic content-based image retrieval: A comprehensive study

Alzu’bi

Amira

Ramzan

2015

Journal of Visual Communication and Image Representation

180

106

View full text Add to dashboard Cite

“…Table II shows that the required memory sizes to store edge-weighted graphs of the Cell-MST method are extremely less than the required memory sizes to store distinct edges of similarity or distance matrixes of the previous similarity-based methods abbreviated to SIM [10,11], and MST-based methods abbreviated to MST [16][17][18]. Table III shows that the initial threshold values 0 and the number of selected edges of the proposed Cell-MST optimal threshold initialization are smaller than the initial threshold values and the number of selected edges of the Prim Trajectory [18].…”

Section: Experiments Resultsmentioning

confidence: 99%

“…Another graph-based K-Means clustering is using minimum spanning trees [16]. A MST-based method has been proposed to be used for centroid initialization associated with cluster number estimation [17].…”

Section: Cluster Number Estimation 1) Without Using Graph Theorymentioning

confidence: 99%

“…Similar to the MST-based methods [16,17], the use of MST combined Prims trajectory for K-means clustering has been suggested [18]. The MST combined Prims trajectory will be constructed from a dataset by using a sorted list of edges which are not larger than the standard deviation of all edge lengths.…”

Section: Cluster Number Estimation 1) Without Using Graph Theorymentioning

confidence: 99%

See 1 more Smart Citation

A cell-MST-based method for big dataset clustering on limited memory computers

Hieu

Meesad

2015

2015 7th International Conference on Information Technology and Electrical Engineering (ICITEE)

View full text Add to dashboard Cite

This paper presents a new clustering algorithm, called Cell-MST-Based Method that is a combination of a Cellbased method and Minimum Spanning Tree based (MST-based) methods. The algorithm is dedicated for Big Datasets on a limited memory computer, especially for thin big datasets which have a small number of attributes but a very large number of instances. Firstly, a Cell-based method converts a big dataset to a small grid of cells in such a way that the required memory to store an edgeweighted graph created from the grid which is less than the available memory of a computer. Then MST-based methods obtain an optimal threshold, estimate the number of clusters and determine the initial centroids. The proposed Cell-MST-based methods can reduce more than 99% of the required memory of the previous similarity-based and MST-based cluster number estimation methods. Moreover, this new Cell-MST-based method also outperforms the quantization error modeling method in terms of executing time and estimated accurate level.

show abstract

“…The time complexity of calculating one delay cost path is O(|V| 2 ) [32]; therefore, the time complexity for a complete graph would be O((|S||D| + |D||D−1| 2 ) × |V| 2 ). The time complexity of forming the minimum spanning tree T is O((|D| + 1) 2 ).…”

Section: Complexity Analysis For Algorithmmentioning

confidence: 99%

Delay Bounded Multi-Source Multicast in Software-Defined Networking

et al. 2018

View full text Add to dashboard Cite

Software-Defined Networking (SDN) is the next generation network architecture with exciting application prospects. The control function in SDN is decoupled from the data forwarding plane, hence it provides a new centralized architecture with flexible network resource management. Although SDN is attracting much attention from both industry and research, its advantage over the traditional networks has not been fully utilized. Multicast is designed to deliver content to multiple destinations. The current traffic engineering in SDN focuses mainly on unicast, however, multicast can effectively reduce network resource consumption by serving multiple clients. This paper studies a novel delay-bounded multi-source multicast SDN problem, in which among the set of potential sources, we select a source to build the multicast-tree, under the constraint that the transmission delay for every destination is bounded. This problem is more difficult than the traditional Steiner minimum tree (SMT) problem, since it needs to find a source from the set of all potential sources. We model the problem as a mixed-integer linear programming (MILP) and prove its NP-Hardness. To solve the problem, a delay bounded multi-source (DBMS) scheme is proposed, which includes a DBMS algorithm to build a minimum delay cost DBMS-Forest. Through a MATLAB experiment, we demonstrate that DBMS is significantly more efficient and outperforms other existing algorithms in the literature.

show abstract

A fast minimum spanning tree algorithm based on K-means

Cited by 93 publications

References 44 publications

Semantic content-based image retrieval: A comprehensive study

Semantic content-based image retrieval: A comprehensive study

A cell-MST-based method for big dataset clustering on limited memory computers

Delay Bounded Multi-Source Multicast in Software-Defined Networking

Contact Info

Product

Resources

About