A fuzzy vector valued KNN-algorithm for automatic outlier detection

Östermark, Ralf

doi:10.1016/j.asoc.2009.03.009

Cited by 23 publications

(10 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Shaikh et al [12] carried out outlier detection on a dataset assumed to obey Gaussian distribution. Östermark [13] put forward the k-th nearest neighbors (k-NN) algorithm, which is applicable to distance-based outlier detection algorithm. Ghoting et al [14] presented the restricted block relocation problem (rBRP) algorithm, making distancebased outlier detection algorithm suitable for higher dimensional datasets.…”

Section: Literature Reviewmentioning

confidence: 99%

A Neighbor Propagation Clustering Algorithm for Intrusion Detection

Li¹

2020

RIA

View full text Add to dashboard Cite

Outlier detection is an important research direction in data mining. This paper mainly applies outlier detection algorithm in intrusion detection, and proposes a novel outlier detection algorithm based on neighbor propagation clustering. The proposed algorithm first clusters the original dataset, then calculates the outlier degree, and finally mines the intrusion behaviors. To verify its effectiveness, our algorithm was tested on public dataset on network intrusions. The results prove that our algorithm can detect intrusions with a high accuracy.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

A Neighbor Propagation Clustering Algorithm for Intrusion Detection

Li¹

2020

RIA

View full text Add to dashboard Cite

show abstract

“…For comparing the results of the proposed framework for reconciliating different final ensemble clusters of different turbines, three other approaches are used, alternatively, to reconciliate the consensus clusters of the FF1 NPP turbine on the basis of the received information from the EE1 NPP turbine that are: 1) clustering of the aggregated shut-down transients of the FF1 and EE1 NPPs turbines by the unsupervised ensemble clustering approach, 2) the inclusion of the EE1 transients into FF1 ensemble clustering by resorting to Fuzzy similarity measure [37][38][39] and 3) the classification of the EE1 transients by a supervised classifier, such as a Fuzzy K-Nearest Neighbours algorithm (FKNN) [40][41][42] trained on FF1 consensus clustering. groups by resorting to the ensemble clustering approach of Section 2, Step 1 (see Appendix A), without any reconciliation (Step 2).…”

Section: Comparison With Other Approachesmentioning

confidence: 99%

“…Several classification algorithms have been proposed and used in practice, like Support Vector Machines (SVM) [52], Naïve Bayes classifier [53], Decision trees [54], Discriminant analysis [55], Classification and Regression Tree (CART) [56,57] and Fuzzy K-Nearest Neighbours (FKNN) [40][41][42]. In this work, we resort to the Fuzzy K-Nearest Neighbors (FKNN) algorithm, because FKNN is simple, requires less computation time during the training phase and is one of the most used [42].…”

Section: Classifying the Shut-down Transients Of The Ee1 Npp Turbine mentioning

confidence: 99%

“…For comparison, three other approaches are used to reconciliate the consensus clusters of the FF1 NPP turbine on the basis of the received information from the EE1 NPP turbine: 1) clustering of the aggregated shut-down transients of FF1 and EE1 NPPs turbines by the unsupervised ensemble clustering approach, 2) the inclusion of the EE1 transients into the FF1 ensemble clustering by resorting to Fuzzy similarity measure [37][38][39] and 3) the classification of EE1 transients by a supervised classifier, such as a Fuzzy K-Nearest Neighbours algorithm (FKNN) [40][41][42] trained on FF1 clustering. Results are discussed and compared with those obtained with the proposed approach: it is concluded that the proposed approach is able to update effectively the clusters of the FF1 NPP turbine on the basis of the received information from the EE1 NPP turbine, and that it is superior to the other approaches with respect to the goodness of the final consensus clustering, computational demand, data requirements, and fault diagnosis effectiveness.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A framework for reconciliating data clusters from a fleet of nuclear power plants turbines for fault diagnosis

Al‐Dahidi

Maio

Baraldi

et al. 2018

Applied Soft Computing

View full text Add to dashboard Cite

When a fleet of similar Systems, Structures and Components (SSCs) is available, the use of all the available information collected on the different SSCs is expected to be beneficial for the diagnosis purpose. Although different SSCs experience different behaviours in different environmental and operational conditions, they maybe informative for the other (even if different) SSCs. In the present work, the objective is to build a fault diagnostic tool aimed at capitalizing the available data (vibration, environmental and operational conditions) and knowledge of a heterogeneous fleet of P Nuclear Power Plants (NPPs) turbines. To this aim, a framework for incrementally learning different clusterings independently obtained for the individual turbines is here proposed. The basic idea is to reconciliate the most similar clusters across the different plants. The data of shutdown transients acquired from the past operation of the P NPPs turbines are summarized into a final, reconciliated consensus clustering of the turbines behaviors under different environmental and operational conditions. Eventually, one can distinguish, among the groups, those of anomalous behavior and relate them to specific root causes. The proposed framework is applied on the shutdown transients of two different NPPs. Three alternative approaches for learning data are applied to the case study and their results are compared to those obtained by the proposed framework: results show that the proposed approach is superior to the other approaches with respect to the goodness of the final consensus clustering, computational demand, data requirements, and fault diagnosis effectiveness.

show abstract

“…EKNN starts from original training set [22]. • Dasarathy [16] has developed a condensing method to edit the reference set: his method provides the minimal consistent subset(MCS) which is used as the editing reference set. All the samples in the reference set can be correctly classified using the k-NN and the MCS [23].…”

Section: B Sample Selectionmentioning

confidence: 99%

A Modified Editing k-nearest Neighbor Rule

Chang¹,

Pei²,

Zhang³

2011

JCP

View full text Add to dashboard Cite

Classification of objects is an important area in a variety of fields and applications. Many different methods are available to make a decision in those cases. The k-nearest neighbor rule (k-NN) is a well-known nonparametric decision procedure. Classification rules based on the k-NN have already been proposed and applied in diverse substantive areas. The editing k-NN proposed by Wilson would be an important one. In this rule, editing the reference set is first performed, every sample in the reference set is classified by using the k-NN rule and the set is formed by eliminating it from the reference set. All the samples mistakenly classified are then deleted from the reference set. Afterward, any input sample is classified using the k-NN rule and the edited reference set. Obviously, the editing k-nearest neighbors classifier (EK -NN) consists of the k-nearest neighbor classifier and an editing reference set. However, the editing reference set gained by this method is only a subset of the reference set. This may result in the loss of some important information and decline of classification accuracy. In this paper, we focus on modifying the editing reference set of EK -NN, the new editing set in our method consists of subsets of the reference set and testing set, such subsets are received by classifying every sample in the reference set and testing set by using the k-NN rule and removing misclassified samples from reference set and testing set, respectively. Advantages of our method are to reduce the loss of information and improve the recognition rate. Comparisons and analysis of the experimental results demonstrate the capability of the proposed algorithm

show abstract

A fuzzy vector valued KNN-algorithm for automatic outlier detection

Cited by 23 publications

References 26 publications

A Neighbor Propagation Clustering Algorithm for Intrusion Detection

A Neighbor Propagation Clustering Algorithm for Intrusion Detection

A framework for reconciliating data clusters from a fleet of nuclear power plants turbines for fault diagnosis

A Modified Editing k-nearest Neighbor Rule

Contact Info

Product

Resources

About