Deep Clustering Analysis via Dual Variational Autoencoder With Spherical Latent Embeddings

被引:21
|
作者
Yang, Lin [1 ]
Fan, Wentao [1 ]
Bouguila, Nizar [2 ]
机构
[1] Huaqiao Univ, Dept Comp Sci & Technol, Xiamen 361021, Peoples R China
[2] Concordia Univ, Concordia Inst Informat Syst Engn CIISE, Montreal, PQ H3G 1T7, Canada
基金
中国国家自然科学基金;
关键词
Clustering; data augmentation; dual variational autoencoder; mixture models; variational autoencoder (VAE); variational inference; von Mises-Fisher mixture model;
D O I
10.1109/TNNLS.2021.3135460
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In recent years, clustering methods based on deep generative models have received great attention in various unsupervised applications, due to their capabilities for learning promising latent embeddings from original data. This article proposes a novel clustering method based on variational autoencoder (VAE) with spherical latent embeddings. The merits of our clustering method can be summarized as follows. First, instead of considering the Gaussian mixture model (GMM) as the prior over latent space as in a variety of existing VAE-based deep clustering methods, the von Mises-Fisher mixture model prior is deployed in our method, leading to spherical latent embeddings that can explicitly control the balance between the capacity of decoder and the utilization of latent embedding in a principled way. Second, a dual VAE structure is leveraged to impose the reconstruction constraint for the latent embedding and its corresponding noise counterpart, which embeds the input data into a hyperspherical latent space for clustering. Third, an augmented loss function is proposed to enhance the robustness of our model, which results in a self-supervised manner through the mutual guidance between the original data and the augmented ones. The effectiveness of the proposed deep generative clustering method is validated through comparisons with state-of-the-art deep clustering methods on benchmark datasets. The source code of the proposed model is available at https://github.com/fwt-team/DSVAE.
引用
收藏
页码:6303 / 6312
页数:10
相关论文
共 50 条
  • [1] Deep clustering analysis via variational autoencoder with Gamma mixture latent embeddings
    Guo, Jiaxun
    Fan, Wentao
    Amayri, Manar
    Bouguila, Nizar
    NEURAL NETWORKS, 2025, 183
  • [2] Deep Clustering With Variational Autoencoder
    Lim, Kart-Leong
    Jiang, Xudong
    Yi, Chenyu
    IEEE SIGNAL PROCESSING LETTERS, 2020, 27 (27) : 231 - 235
  • [3] Deep Clustering of Compressed Variational Embeddings
    Wu, Suya
    Diao, Enmao
    Ding, Jie
    Tarokh, Vahid
    2020 DATA COMPRESSION CONFERENCE (DCC 2020), 2020, : 399 - 399
  • [4] scBGEDA: deep single-cell clustering analysis via a dual denoising autoencoder with bipartite graph ensemble clustering
    Wang, Yunhe
    Yu, Zhuohan
    Li, Shaochuan
    Bian, Chuang
    Liang, Yanchun
    Wong, Ka-Chun
    Li, Xiangtao
    BIOINFORMATICS, 2023, 39 (02)
  • [5] Deep Clustering with Spherical Distance in Latent Space
    Tran, Bach
    Le Thi, Hoai An
    ADVANCED COMPUTATIONAL METHODS FOR KNOWLEDGE ENGINEERING (ICCSAMA 2019), 2020, 1121 : 231 - 242
  • [6] Unsupervised Seismic Facies Deep Clustering Via Lognormal Mixture-Based Variational Autoencoder
    Hua, Haowei
    Qian, Feng
    Zhang, Gulan
    Yue, Yuehua
    IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2023, 16 : 9831 - 9842
  • [7] Deep Spectral Clustering using Dual Autoencoder Network
    Yang, Xu
    Deng, Cheng
    Zheng, Feng
    Yan, Junchi
    Liu, Wei
    2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019), 2019, : 4061 - 4070
  • [8] Unsupervised convolutional variational autoencoder deep embedding clustering for Raman spectra
    Guo, Yixin
    Jin, Weiqi
    Wang, Weilin
    Guo, Zongyu
    He, Yuqing
    ANALYTICAL METHODS, 2022, 14 (39) : 3898 - 3910
  • [9] Deep clustering of small molecules at large-scale via variational autoencoder embedding and K-means
    Hamid Hadipour
    Chengyou Liu
    Rebecca Davis
    Silvia T. Cardona
    Pingzhao Hu
    BMC Bioinformatics, 23
  • [10] Deep clustering of small molecules at large-scale via variational autoencoder embedding and K-means
    Hadipour, Hamid
    Liu, Chengyou
    Davis, Rebecca
    Cardona, Silvia T.
    Hu, Pingzhao
    BMC BIOINFORMATICS, 2022, 23 (SUPPL 4)