Bayesian distance metric learning for discriminative fuzzy c-means clustering

Heidari, Negar; Moslehi, Zahra; Mirzaei, Abdolreza; Safayani, Mehran

doi:10.1016/j.neucom.2018.08.071

Cited by 12 publications

(12 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Learning a Mahalanobis distance metric designs a new distance measurement function that can learn the Mahalanobis distance metric by forcibly adjusting the distance of a given instance and applying it to new data [19]. Bayesian discriminative fuzzy clustering (BDFC) designs a probabilistic method for unsupervised distance metric learning which can maximize the separability between different clusters in the projection space [20]. e above methods can be regarded as the clustering algorithm based on the distance metric learning.…”

Section: Related Workmentioning

confidence: 99%

Semisupervised Deep Embedded Clustering with Adaptive Labels

Chen

Gao

et al. 2021

Scientific Programming

View full text Add to dashboard Cite

Deep embedding clustering (DEC) attracts much attention due to its outperforming performance attributed to the end-to-end clustering. However, DEC cannot make use of small amount of a priori knowledge contained in data of increasing volume. To tackle this challenge, a semisupervised deep embedded clustering algorithm with adaptive labels is proposed to cluster those data in a semisupervised end-to-end manner on the basis of a little priori knowledge. Specifically, a deep semisupervised clustering network is designed based on the autoencoder paradigm and deep clustering, which well mine the clustering representation and clustering assignment by preventing the shift of labels in DEC. Then, to train parameters of the deep semisupervised clustering network, a back-propagation-based algorithm with adaptive labels is introduced based on the pretrain and fine-tune strategies. Finally, extensive experiments on representative datasets are conducted to evaluate the performance of the proposed method in terms of clustering accuracy and normalized mutual information. Results show the proposed method outperforms the state-of-the-art methods of DEC.

show abstract

Section: Related Workmentioning

confidence: 99%

Semisupervised Deep Embedded Clustering with Adaptive Labels

Chen

Gao

et al. 2021

Scientific Programming

View full text Add to dashboard Cite

show abstract

“…If there are only a small amount of labeled data available, there are no reliable models to select, so it is difficult to determine the best parameters for measurement of similarity based on prior knowledge [11,12]. In recent years, the representation based on sparse coding has attracted much attention [6][7][8]. In order to obtain flexible similarity matrix and get better performance, Wang et al [13] utilized sparse representation to learn similarity, and proposed similarity learning based on sparse representation for semi-supervised boosting.…”

Section: Related Workmentioning

confidence: 99%

“…Recently, in the adaptive field of data similarity, representation based on sparse coding has received extensive attention [6][7][8]. Therefore, this paper uses sparse representation as a measure of similarity.…”

Section: Introductionmentioning

confidence: 99%

Semi-Supervised Boosting Using Similarity Learning Based on Modular Sparse Representation With Marginal Representation Learning of Graph Structure Self-Adaptive

Gao

2020

IEEE Access

View full text Add to dashboard Cite

The purpose of semi-supervised boosting strategy is to improve the classification performance of one given classifier for a large number of unlabeled data. In the semi-supervised boosting strategy, the unlabeled samples are assigned for pseudo labels according to similarities between the labeled samples and the unlabeled samples, and the unlabeled samples with high confidences of pseudo labels are selected as labeled samples at the same time. Good similarities help to assign more appropriate pseudo labels to the unlabeled samples. These selected samples with pseudo labels will be used as the labeled samples to train the new ensemble classifier. Therefore, good and distinguishable similarities learning between unlabeled samples and labeled samples has shown remarkable importance due to its promising performance for semi-supervised boosting strategy. This paper presents semi-supervised boosting framework using similarity learning based on modular sparse representation by employing a marginal regression function with probabilistic graphical structure adaptation. In this paper, distinguishable regression targets analysis, graph structure adaptation, robust modular sparse representation and semi-supervised boosting learning are seamlessly incorporated into a joint framework. This framework learns marginal regression targets from data rather than exploiting the conventional zero-one matrix that greatly hinders the freedom of regression fitness and degrades the performance of regression results to improve the interclass separation of the learned representation. Meanwhile, a regularization term based on probabilistic connection knowledge is used to construct a graph regularization with adaptive optimization, which improves the intra-class compactness of the learned representation. Additionally, modular sparse representation learning is used to improve the robustness of the learned representation. The experimental results on four datasets including face and object show that the recognition rates of the proposed method are significantly better than other state-of-the-art methods.

show abstract

“…Klasterisasi merupakan metode yang paling umum digunakan untuk mengelompokkan objek berdasarkan kemiripan dengan objek yang lain [1]. Terdapat beberapa pendekatan metode klasterisasi diantaranya Partitioning Methods [2] [3], Hierarchical Methods [4] [5], Density Based Methods [6] [7], Grid Based Methods [8], dan Model Based Clustering Methods [9]. Metode klasterisasi yang dipilih sangat bergantung pada permasalahan yang akan diselesaikan.…”

Section: Pendahuluanunclassified

Klasterisasi Dokumen Menggunakan Weighted K-Means Berdasarkan Relevansi Topik

Riduwan

Fatichah

Yuniarti

2019

JUTI

View full text Add to dashboard Cite

3) ABSTRAK Jumlah penelitian di dunia mengalami perkembangan yang pesat, setiap tahun berbagai peneliti dari penjuru dunia menghasilkan publikasi ilmiah seperti makalah, jurnal, buku dsb. Metode klasterisasi dapat digunakan untuk mengelompokkan kumpulan dokumen publikasi ilmiah ke dalam suatu kelompok tertentu berdasarkan relevansi antar topik. Klasterisasi pada dokumen memiliki karakteristik yang berbeda karena tingkat kemiripan antar dokumen dipengaruhi oleh kata-kata pembentuknya. Beberapa metode klasterisasi kurang memperhatikan nilai semantik dari kata. Sehingga klaster yang terbentuk kurang merepresentasikan isi topik dokumen. Klasterisasi dokumen teks masih memiliki kemungkinan adanya outlier karena pemilihan fitur teks yang tidak optimal. Oleh karena itu dibutuhkan pemrosesan data yang tepat serta metode yang mengoptimalkan hasil klaster. Penelitian ini mengusulkan metode klasterisasi dokumen menggunakan Weighted K-Means yang dipadukan dengan Maximum Common Subgraph. Weighted k-means digunakan untuk klasterisasi awal dokumen berdasarkan kata-kata yang diekstraksi. Pembentukan Weighted K-Means berdasarkan perhitungan Word2Vec dan TextRank dari katakata dalam dokumen. Maximum Common Subgraph merupakan tahap pembentukan graf yang digunakan dalam penggabungan klaster untuk menghasilkan klaster baru yang lebih optimal. Pembentukan graf dilakukan dengan perhitungan nilai Word2vec dan Co-occurrence dari klaster. Representasi topik dokumen tiap klaster dapat dihasilkan dari pemodelan topik Latent Dirichlet Allocation (LDA). Pengujian dilakukan dengan menggunakan dataset publikasi ilmiah dari Scopus. Hasil dari analisis Koherensi topik menunjukkan nilai koherensi usulan metode adalah 0,5375 pada dataset 1 yang bersifat heterogen dan 0,5642 pada dataset 2 yang bersifat homogen.

show abstract

Bayesian distance metric learning for discriminative fuzzy c-means clustering

Cited by 12 publications

References 12 publications

Semisupervised Deep Embedded Clustering with Adaptive Labels

Semisupervised Deep Embedded Clustering with Adaptive Labels

Semi-Supervised Boosting Using Similarity Learning Based on Modular Sparse Representation With Marginal Representation Learning of Graph Structure Self-Adaptive

Klasterisasi Dokumen Menggunakan Weighted K-Means Berdasarkan Relevansi Topik

Contact Info

Product

Resources

About