Cluster validation using an ensemble of supervised classifiers

Rodríguez, Jorge; Medina‐Pérez, Miguel Angel; Gutierrez-Rodríguez, Andres E.; Monroy, Raúl; Terashima-Marín, Hugo

doi:10.1016/j.knosys.2018.01.010

Cited by 25 publications

(33 citation statements)

References 64 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Not all internal indexes measure compactness and separation (Riyaz & Wani, 2016). In our experimentation, we choose VIC (Rodríguez et al, 2018), an internal index whose design is not focused on the shape of the clusters but on the rationale that a good partition also yields a good classification model.…”

Section: Internal External and Relative Indexesmentioning

confidence: 99%

“…VIC was tested using 50 different data sets, where it significantly outperforms other well known cluster indexes (Rodríguez et al, 2018). Unlike other internal indexes, which tend to prefer clusters with specific shapes, such as hyperspheres (Lago-Fernández & Corbacho, 2010;Halkidi & Vazirgiannis, 2008;Gurrutxaga et al, 2011), VIC does not assume a specific shape.…”

Section: Validity Index Using Supervised Classifiersmentioning

confidence: 99%

“…First, we present the purpose of cluster validity indexes. Later, we introduce validity index using supervised classifiers (VIC; Rodríguez et al, ), the cluster validity index used by our proposed classifier, OCKRA++.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Cluster validation in clustering‐based one‐class classification

Rodríguez¹,

Monroy²,

Medina‐Pérez³

et al. 2019

Expert Systems

Self Cite

View full text Add to dashboard Cite

Reconstruction‐based one‐class classification has shown to be very effective in a number of domains. This approach works by attempting to capture the underlying structure of the normal class, typically, by means of clusters of objects. It has the main disadvantage, however, that one has to indicate the number of clusters in advance, for this yields an efficient way of computing a clustering. In this paper, we introduce a new algorithm, OCKRA++, which achieves a better performance, by enhancing a clustering‐based one‐class ensemble classifier (OCKRA) with a cluster validity index that is used to set the best number of clusters during the classifier's training process. We have thoroughly tested OCKRA++ in a particular domain, namely masquerade detection. For this purpose, we have used the Windows‐Users and ‐Intruder simulation Logs data set repository, which contains 70 different masquerade data sets. We have found that OCKRA++ is currently the algorithm that achieves the best area under the curve, with a significant difference, in masquerade detection using the file system navigation approach.

show abstract

Section: Internal External and Relative Indexesmentioning

confidence: 99%

Section: Validity Index Using Supervised Classifiersmentioning

confidence: 99%

See 1 more Smart Citation

Cluster validation in clustering‐based one‐class classification

Rodríguez¹,

Monroy²,

Medina‐Pérez³

et al. 2019

Expert Systems

Self Cite

View full text Add to dashboard Cite

show abstract

“…It is difficult to select the cluster member; you need time to adapt to changes and to prevent unpredictable situations [4]. Also, some authors [5] propose to use the index of the cluster reliability used to select the clustering algorithm of the enterprise. It works by assessing the quality of the members of the cluster formation, the formation of the major candidates in the cluster, and evaluate their skills.…”

Section: Main Bodymentioning

confidence: 99%

Effective Cluster Model of Thermal Power Company Management

Chevhanova¹,

Chichulina²,

Skryl³

2018

IJET

View full text Add to dashboard Cite

The aim of this study is to determine the need for cluster formations; to identify the influence factors on the creation of a cluster of thermal power companies in Poltava region. The method of evaluating the efficiency of formation of this cluster has been developed. The methodological framework of this research is general scientific dialectical method of cognition, in which the research object is studied as a dynamic system in the process of its development. The formation of the main directions of thermal power company development was carried out on the basis of dialectical, historical and systematic methods. In the process of studying and generalization of scientific and practical development methods of comparison, analysis and synthesis, induction and deduction were applied. Also the study is based on regulatory and economic instruments, economical studies and studies of scientific research institutions. As a result of the study the main participants of the cluster formation were identified. So it may be comprised of the members of the cluster, generating thermal energy, the thermal energy consumers and the region where the cluster of thermal power companies will be formed. According to the participants of cluster associations the factors of influence on economic and social effect from implementation of the cluster thermal power companies in the Poltava region were formed. We determined that to reduce the rate for thermal energy it is necessary to conduct a number of activities. The application of the proposed measures will significantly reduce the services cost of thermal power companies. The cost of thermal energy and the distribution among the participants of cluster associations were calculated on the example of Poltava region utility production enterprise of heat economy "Poltavateploenergo". The participants are heat producers, transport companies, distribution and heat supply companies, repair work companies, construction work enterprises. It is established that the overall effect of cluster members has both quantitative and qualitative nature. The impact on consumers is qualitative. It is based on the fact that the tariff for consumers remained unchanged in the medium term (five years). The essence of economic effect is in the quality and timeliness of services thermal management. The overall effect of its work influences all members. It occurs when there is the formation and implementation of cluster thermal power companies. But in addition to the advantages of cluster associations its major problems were presented in the work. Among them there are the following problems: the lack of informativeness of executive authorities and business representatives in the application of the cluster approach; the lack of public policy to ensure the systematic approach and organization of interaction between different levels of executive authorities in the implementation of cluster projects; lack of institutional and financial support to cluster initiatives; lack of trained personnel in the organizational aspects o...

show abstract

“…The SOM architecture consists of an input layer with n training vector units, output layer with k category / cluster and intra-layer unit that connects between input layer and output layer [7,8] …”

Section: Self-organizing Maps (Som)mentioning

confidence: 99%

Artificial Neural Network Optimized Approach for Improving Spatial Cluster Quality of Land Value Zone

Haviluddin¹,

Agus²,

Azhari³

et al. 2018

IJET

View full text Add to dashboard Cite

A geostatistics practical approach is divided data sample into several groups with certain rules. Then, the data groups are used for spatial interpolation. Furthermore, clustering technique is quite commonly used in order to get distance function between sample data. In this study, Self-Organizing Maps (SOM) optimized by using Learning Vector Quantization (LVQ) especially in distance variance have been implemented. The land value zone datasets in Samarinda, East Kalimantan, Indonesia have been used. This study shows that the SOM optimized by LVQ technique have a good distance variance value in the same cluster than SOM technique. In other words, SOM-LVQ can be alternative clustering technique especially centroid position in clusters.

show abstract

Cluster validation using an ensemble of supervised classifiers

Cited by 25 publications

References 64 publications

Cluster validation in clustering‐based one‐class classification

Cluster validation in clustering‐based one‐class classification

Effective Cluster Model of Thermal Power Company Management

Artificial Neural Network Optimized Approach for Improving Spatial Cluster Quality of Land Value Zone

Contact Info

Product

Resources

About