Validity of internal cluster indices

Vij, Ankit; Khandnor, Padmavati

doi:10.1109/csitss.2016.7779392

Cited by 4 publications

(3 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In [18], the authors proposed singular value decomposition to extract features before k-means clustering and evaluate the error sum of squares (SSE) index to compare with direct clustering. In [19], they used a fused load curve k-means algorithm, based on "Haar" discrete wavelet transform for reduce dimension, to obtain the load patterns of consumers from China and the United States and evaluate clustering performance by four CVI [20]. Xiao et al [21] proposed a fusion clustering algorithm to obtain the consumption characteristics, using load curve clustering, based on discrete wavelet transform (CC-DWT).…”

Section: Related Workmentioning

confidence: 99%

Electricity Pattern Analysis by Clustering Domestic Load Profiles Using Discrete Wavelet Transform

2022

View full text Add to dashboard Cite

Energy demand has grown explosively in recent years, leading to increased attention of energy efficiency (EE) research. Demand response (DR) programs were designed to help power management entities meet energy balance and change end-user electricity usage. Advanced real-time meters (RTM) collect a large amount of fine-granular electric consumption data, which contain valuable information. Understanding the energy consumption patterns for different end users can support demand side management (DSM). This study proposed clustering algorithms to segment consumers and obtain the representative load patterns based on diurnal load profiles. First, the proposed method uses discrete wavelet transform (DWT) to extract features from daily electricity consumption data. Second, the extracted features are reconstructed using a statistical method, combined with Pearson’s correlation coefficient and principal component analysis (PCA) for dimensionality reduction. Lastly, three clustering algorithms are employed to segment daily load curves and select the most appropriate algorithm. We experimented our method on the Manhattan dataset and the results indicated that clustering algorithms, combined with discrete wavelet transform, improve the clustering performance. Additionally, we discussed the clustering result and load pattern analysis of the dataset with respect to the electricity pattern.

show abstract

Section: Related Workmentioning

confidence: 99%

Electricity Pattern Analysis by Clustering Domestic Load Profiles Using Discrete Wavelet Transform

2022

View full text Add to dashboard Cite

show abstract

“…Beberapa parameter yang menjadi karakteristik adalah kerapatan pemisahan yang diukur pada jarak intra-cluster [14]. Dimana metode pengukuran validitas cluster dapat dikategorikan menjadi dua yaitu external dan internal [5], akan tetapi ada juga yang mengkategorikannya menjadi tiga yaitu validasi internal, external, dan relatif [15]. Silhouette berguna ketika mencari cluster yang rapat dan benar-benar terpisah.…”

Section: Metode Pengukuran Validitas Clusterunclassified

Penentuant Jumlah Cluster Optimum Pada Segmen Rute Penerbangan Menggunakan Data Automatic Dependent Surveillance-Broadcast

Prayogo

Buliali

2020

JUTI

View full text Add to dashboard Cite

2) ABSTRAK Terdapat beberapa titik acuan dalam satu rute penerbangan untuk keperluan navigasi yang disebut waypoint. Pada penelitian ini penulis melakukan segmentasi untuk membagi satu rute penerbangan (Surabaya-Palu) menjadi 7 segmen yang terdiri dari 8 waypoint, dengan membuat garis imajiner secara tegak lurus melewati masing-masing waypoint. Pada tiap segmen dilakukan analisa terkait lokasi yang paling sering dilalui menggunakan pendekatan clustering. Dalam penelitian ini penulis menggunakan algoritma clustering K-means dengan optimasi centroid yang mengimplementasikan algoritma Ant Lion Optimizer (ALO) atau disebut dengan K-means-ALO. Jumlah cluster ditentukan sebelumnya, kemudian dilakukan validasi pengelompokan internal dengan menggunakan silhouette index. Hasil metode pengelompokan diuji nilai performansinya. Hasil akhir dari jumlah cluster yang sudah ditentukan diambil nilai validitas cluster terbaik yaitu jumlah cluster yang optimum pada tiap segmen area penerbangan. Pengujian dilakukan dengan membandingkan nilai silhouette index untuk setiap percobaan jumlah cluster terhadap kedua metode yaitu K-means dan K-means-ALO. Pada uji coba yang dilakukan, metode optimasi yang diusulkan menghasilkan validitas cluster yang lebih baik sesuai nilai silhouette index pada tiga segmen, yaitu segmen 2, 3, dan 5 akan tetapi signifikan di semua segmen berdasarkan uji statistik Analysis of Variance (ANOVA) dan uji lanjut Least Significant Difference (LSD).

show abstract

“…More often than not, such indices are designed to be "the best" for a given particular situation and/or they aim to "eliminate" certain deficiencies with the previous measures. Because of this, the number of CVIs to choose from can be overwhelming, see Section 2 and, e.g., [56,58], for up-to-date overviews.…”

Section: Introductionmentioning

confidence: 99%

Are Cluster Validity Measures (In)valid?

Gagolewski,

Bartoszuk,

Cena

2022

Preprint

View full text Add to dashboard Cite

Internal cluster validity measures (such as the Caliński-Harabasz, Dunn, or Davies-Bouldin indices) are frequently used for selecting the appropriate number of partitions a dataset should be split into. In this paper we consider what happens if we treat such indices as objective functions in unsupervised learning activities. Is the optimal grouping with regards to, say, the Silhouette index really meaningful? It turns out that many cluster (in)validity indices promote clusterings that match expert knowledge quite poorly. We also introduce a new, well-performing variant of the Dunn index that is built upon OWA operators and the near-neighbour graph so that subspaces of higher density, regardless of their shapes, can be separated from each other better.

show abstract

Validity of internal cluster indices

Cited by 4 publications

References 29 publications

Electricity Pattern Analysis by Clustering Domestic Load Profiles Using Discrete Wavelet Transform

Electricity Pattern Analysis by Clustering Domestic Load Profiles Using Discrete Wavelet Transform

Penentuant Jumlah Cluster Optimum Pada Segmen Rute Penerbangan Menggunakan Data Automatic Dependent Surveillance-Broadcast

Are Cluster Validity Measures (In)valid?

Contact Info

Product

Resources

About