Abstract:Latent low-rank representation (LLRR) is a critical method to uncover the low-dimensional subspace structure embedded in high-dimensional data, especially when the data sampling is insufficient and/or extremely corrupted. However, LLRR uses the nuclear norm as a convex relaxation of the rank function, and thus the rank may not be well approximated in practice since the nuclear norm treats all singular values equally. To overcome this deficiency, in this paper, we propose a new model, TS1-LLRR, which extends LLRR along with its corresponding subspace clustering algorithm. We first adopt the non-convex transformed Schatten-1 (TS1) function, a better approximation to the rank of the matrix than the nuclear norm, to model the coefficient matrix. The proposed model works well in learning intrinsic rank information and thus has a stronger ability to capture the subspace structure. Furthermore, to solve the proposed non-convex model, we design an efficient iterative algorithm by using the alternating direction method of multipliers (ADMM) framework. This algorithm performs well in both accuracy and convergence speed, and the resulting subproblems have closed-form solutions. Last, we construct the affinity matrix for spectral clustering by further exploiting the angular information of the principal directions of the low-rank representation, which enhances the capability of investigating the inherent correlation between data points. Extensive experimental results on image clustering and motion segmentation show that our proposed subspace clustering method outperforms the compared methods.

STMLRC: Sparse Topic Model with Low Rank Constraint

Efficient Probabilistic Latent Semantic Analysis with Sparsity Control

Sparse online topic models

Probabilistic Latent Semantic Analysis for Sketch-Based 3D Model Retrieval

A Hierarchical Latent Topic Model Based on Sparse Coding

Parsimonious Topic Models with Salient Word Discovery

Mitigating Data Sparsity for Short Text Topic Modeling by Topic-Semantic Contrastive Learning

Sparse Relational Topic Models for Document Networks.

Group Sparse Topical Coding

Sparse Relational Topical Coding on Multi-Modal Data.

Joint Topic-Document Modeling Via Low-Dimensional Sparse Models

Spatial Topic Pyramid Model: Topic Model with Regional Spatial Information.

Locally discriminative topic modeling

Refine Bigram PLSA Model by Assigning Latent Topics Unevenly

Efficient Topic Modeling on Phrases via Sparsity

Efficient Methods for Incorporating Knowledge into Topic Models

Topic-weak-correlated Latent Dirichlet Allocation

Subspace Clustering Based on Latent Low-Rank Representation with Transformed Schatten-1 Penalty Function

Robust sparse low-rank embedding for image dimension reduction

Conditional topical coding: an efficient topic model conditioned on rich features.

A Latent Topic Model for Linked Documents