Abstract:Clustering algorithm is one of the most widely used and influential analysis techniques. With the advent of deep learning, deep embedding clustering algorithms have rapidly evolved and yield promising results. Much of the success of these algorithms depends on the potential expression captured by the autoencoder network. Therefore, the quality of the potential expression directly determines the algorithm's performance. In view of this, researchers have proposed many improvements. Although the performance has been slightly improved, they all have one shortcoming, that is, too much emphasis is placed on the original data reconstruction ability during the process of feature expression, which greatly limits the further expression of potential features according to specific clustering tasks. Moreover, there is a large amount of noise in the original data, so blindly emphasizing reconstruction will only backfire. Hence, we innovatively propose a deep embedding clustering algorithm based on residual autoencoder (DECRA) after in-depth research. Specifically, a novel autoencoder network with residual structure is proposed and introduced into deep embedded clustering tasks. The network introduces an adaptive weight layer in feature representation z, which can make it have good robustness, generalization for specific tasks, and adaptive learning of better feature embeddings according to category classification. In this paper, the reasons for the validity of this structure are explained theoretically, and comprehensive experiments on six benchmark datasets including various types show that the clustering performance of the DECRA is very competitive and significantly superior to the most advanced methods.

Autoencoder Constrained Clustering with Adaptive Neighbors.

Nonlinear Subspace Clustering Via Adaptive Graph Regularized Autoencoder.

Adversarially Regularized Joint Structured Clustering Network

Deep Embedded Clustering with Asymmetric Residual Autoencoder

Structure Enhanced Deep Clustering Network Via a Weighted Neighbourhood Auto-Encoder

Structure-Constrained Feature Extraction by Autoencoders for Subspace Clustering

SGAE: Stacked Graph Autoencoder for Deep Clustering

Subspace Clustering Via Integrating Sparse Representation and Adaptive Graph Learning.

Learning with Adaptive Neighbors for Image Clustering

Adaptive Graph Auto-Encoder for General Data Clustering

Deep clustering based on embedded auto-encoder

Graph Embedding Clustering: Graph Attention Auto-Encoder with Cluster-Specificity Distribution.

Deep Clustering with Convolutional Autoencoders

Attributed Graph Clustering Network with Adaptive Feature Fusion

Deep Embedding Clustering Based on Residual Autoencoder

Adaptive Graph Convolutional Clustering Network with Optimal Probabilistic Graph

Stacked Sparse Auto-Encoder for Deep Clustering.

Representation Learning Based on Autoencoder and Deep Adaptive Clustering for Image Clustering

Clustering with Orthogonal AutoEncoder

Deep self-supervised clustering with embedding adjacent graph features

Subspace Clustering Using a Low-Rank Constrained Autoencoder