The Initialization of Flexible K-Medoids Partitioning Method Using a Combination of Deviation and Sum of Variable Values

Kariyam, Kariyam; Abdurakhman, Abdurakhman; Subanar, Subanar; Utami, Herni

doi:10.13189/ms.2022.100501

Cited by 3 publications

(6 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The transformation for the non-missing ordinal, interval, and ratio scale, is in two steps below [8]:…”

Section: Transformation Methods and Proximity Measurementioning

confidence: 99%

“…For the partitioning method, we use k-means [20], simple and fast k-medoids [5], and modification of flexible k-medoids (proposed method). Reference [8] shows that a representative object of the block of the combined standard deviation and the sum of variable values as the initial medoids guarantees no empty initial groups.…”

Section: Partitioning Methods (Proposed)mentioning

confidence: 99%

“…Suppose we have n objects with p-variables of numerical or categorical or mixed, then the standard deviation for an object 𝑖 with 𝑝-variables is as follows [8]:…”

Section: Partitioning Methods (Proposed)mentioning

confidence: 99%

“…When medoids are non-unique objects, the SFKM algorithm suffers from possible empty clusters, and in the SKM algorithm, similar things may be in different groups. The first phase of the flexible k-medoids (FKM) ensures that no initial groups are empty and the medoids of identical objects are in the same groups [8]. Another issue in k-medoids clustering is the unpredictable number of iterations.…”

Section: Introductionmentioning

confidence: 99%

“…In contrast, the SKM method suggests twenty times initialization [7]. Like the SKM, the FKM algorithm randomly implements five to fifteen times to update the medoid based on the initial group members [8]. The purity algorithm uses the Davies-Bouldin Index to analyze groups to reduce the number of iterations in k-medoids [10], making this method suitable for numerical data.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Block-Based K-Medoids Partitioning Method with Standardized Data to Improve Clustering Accuracy

Kariyam¹,

Abdurakhman²,

Subanar³

et al. 2022

MMEP

View full text Add to dashboard Cite

Most of the existing k-medoid algorithms select the initial medoid randomly or use a specific formula based on the proximity matrix. This study proposes a block-based k-medoids partitioning method for clustering objects. To get the initial medoids, we search for an object representative from the block of the standard deviation and the sum of the variable values. We optimized the initial groups to update medoids, so this step can reduce the number of iterations to obtain partitioned data. The block-based k-medoids partitioning method applies to all types of data. To improve clustering accuracy, we operate pre-processing through data standardization. We conducted a series of experiments on eight real data sets and three artificial data to evaluate the proposed method's performance in terms of clustering accuracy. The experiment results show that the Block-based K-Medoids partitioning is more efficient in reducing the number of iterations. The clustering accuracy of the Block-KM for eight real datasets is also comparable to other methods. The data standardization is effective to increase clustering accuracy, especially for block k-medoids, k-means, simple and fast k-medoids, and the Ward method.

show abstract

“…The transformation for the non-missing ordinal, interval, and ratio scale, is in two steps below [8]:…”

Section: Transformation Methods and Proximity Measurementioning

confidence: 99%

Section: Partitioning Methods (Proposed)mentioning

confidence: 99%

“…Suppose we have n objects with p-variables of numerical or categorical or mixed, then the standard deviation for an object 𝑖 with 𝑝-variables is as follows [8]:…”

Section: Partitioning Methods (Proposed)mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%