Clustering with Deep Learning: Taxonomy and New Methods

Abstract

Clustering methods based on deep neural networks have proven promising for clustering real-world data because of their high representational power. In this paper, we propose a systematic taxonomy of clustering methods that utilize deep neural networks. We base our taxonomy on a comprehensive review of recent work and validate the taxonomy in a case study. In this case study, we show that the taxonomy enables researchers and practitioners to systematically create new clustering methods by selectively recombining and replacing distinct aspects of previous methods with the goal of overcoming their individual limitations. The experimental evaluation confirms this and shows that the method created for the case study achieves state-of-the-art clustering quality and surpasses it in some cases.

References in corpus (4)

• Learning Discrete Representations via Information Maximizing Self-Augmented Training
• Deep Learning with Nonparametric Clustering
• Deep Clustering via Joint Convolutional Autoencoder Embedding and Relative Entropy Minimization
• Discriminatively Boosted Image Clustering with Fully Convolutional Auto-Encoders

Cited by in corpus (36)

• Survey of state-of-the-art mixed data clustering algorithms
• Deep Transparent Prediction through Latent Representation Analysis
• Physics-informed cluster analysis and a priori efficiency criterion for the construction of local reduced-order bases
• Centroid Transformers: Learning to Abstract with Attention
• Large-Scale Hyperspectral Image Clustering Using Contrastive Learning
• Improving Image Clustering With Multiple Pretrained CNN Feature Extractors
• Deep Amortized Clustering
• Centroid Based Concept Learning for RGB-D Indoor Scene Classification
• Memory-Based Graph Networks
• CARL-G: Clustering-Accelerated Representation Learning on Graphs
• Cluster-Guided Unsupervised Domain Adaptation for Deep Speaker Embedding
• Neural Clustering Processes
• Unsupervised Embedding Learning for Human Activity Recognition Using Wearable Sensor Data
• Deep Learning and Traffic Classification: Lessons learned from a commercial-grade dataset with hundreds of encrypted and zero-day applications
• Multi-Facet Clustering Variational Autoencoders
• Dense Video Captioning Using Unsupervised Semantic Information
• Meta-Learning to Cluster
• Exploiting Contextual Information with Deep Neural Networks
• Unsupervised Embedding of Hierarchical Structure in Euclidean Space
• Thinkback: Task-SpecificOut-of-Distribution Detection
• Dissimilarity Mixture Autoencoder for Deep Clustering
• Deep Inverse Feature Learning: A Representation Learning of Error
• Learning to Structure an Image with Few Colors
• Amortized Probabilistic Detection of Communities in Graphs
• GMM-Based Generative Adversarial Encoder Learning
• A Framework for Deep Constrained Clustering
• Sparse Label Smoothing Regularization for Person Re-Identification
• Signal Clustering with Class-independent Segmentation
• Multi-level Feature Learning on Embedding Layer of Convolutional Autoencoders and Deep Inverse Feature Learning for Image Clustering
• Representation Learning for Short Text Clustering
• Superpixel Sampling Networks
• Label-Removed Generative Adversarial Networks Incorporating with K-Means
• The Deep Latent Position Topic Model for Clustering and Representation of Networks with Textual Edges
• From Time Series to Euclidean Spaces: On Spatial Transformations for Temporal Clustering
• Quantization-Based Regularization for Autoencoders
• Famous Companies Use More Letters in Logo:A Large-Scale Analysis of Text Area in Logo