Search

Multiple-instance learning (MIL) is a weakly supervised learning paradigm in which a training dataset contains a set of labeled bags, and each bag contains multiple number of unlabeled instances. Preparation of instance-level labels is resource intensive. Being weakly supervised, MIL is sensitive to several practical issues such as noisy label information and low witness rate. A scenario of high class imbalance and low degree-of-supervision further poses additional challenges. Recent works on graph-based label propagation methods have been shown to be effective in semi-supervised setup to address such issues by propagating the label information over graph-based manifold. Application of such semi-supervised strategies for MIL framework requires the instance-level labeling. Whenever the problem setup contains the three characteristics of high class imbalance, low degree-of-supervision and weak supervision, the state-of-the-art methods of either MIL or graph-based label propagation are inadequate when applied alone. This article proposes a non-convex formulation for instance-level MIL to find the instance-level labels by combining the benefits of both MIL and graph-based label propagation methods. The proposed approach improves the performance of the classifier using density-difference- and distance-based structural smoothness assumptions in the graph structure. This article presents the comparison of the performance of the proposed method to those of several state-of-the-art base-lines in MIL. The experimental results are shown on multiple datasets from four different applications. The proposed method is compared in a total of 616 cases (14 datasets × 11 base-line models × 4 low degree-of-supervision values). The minimum f-score improvements are 15.22%, 1.14%, and 4.25% in DAP, CIR, and ACSV datasets, respectively. [ABSTRACT FROM AUTHOR]

Traditional graph-based Semi-Supervised Learning (SSL) methods usually contain two separate steps. First, constructing an affinity matrix. Second, inferring the unknown labels. While such a two-step method has been successful, it cannot take full advantage of the correlation between affinity matrix and label information. In order to address the above problem, we propose a novel graph-based SSL method. It can learn the affinity matrix and infer the unknown labels simultaneously. Moreover, feature selection with auto-weighting is introduced to extract the effective and robust features. Further, the proposed method learns the data similarity matrix by assigning the adaptive neighbors for each data point based on the local distance. We solve the unified problem via an alternative minimization algorithm. Extensive experimental results on synthetic data and benchmark data show that the proposed method consistently outperforms the state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

• A constrained sparse graph construction method is proposed. • The method does not require a predefined affinity matrix. • The proposed constraints impose edge weights smoothness. • The proposed constraints lead to a structured sparse graph. • Performance is assessed on graph-based label propagation and embedding. Sparse representation is a useful tool in machine learning and pattern recognition area. Sparse graphs (graphs constructed using sparse representation of data) proved to be very informative graphs for many learning tasks such as label propagation, embedding, and clustering. It has been shown that constructing an informative graph is one of the most important steps since it significantly affects the final performance of the post graph-based learning algorithm. In this paper, we introduce a new sparse graph construction method that integrates manifold constraints on the unknown sparse codes as a graph regularizer. These constraints seem to be a natural regularizer that was discarded in existing state-of-the art graph construction methods. This regularizer imposes constraints on the graph coefficients in the same way a locality preserving constraint imposes on data projection in non-linear manifold learning. The proposed method is termed Sparse Graph with Laplacian Smoothness (SGLS). We also propose a kernelized version of the SGLS method. A series of experimental results on several public image datasets show that the proposed methods can out-perform many state-of-the-art methods for the tasks of label propagation, nonlinear and linear embedding. [ABSTRACT FROM AUTHOR]

Abstract Data-driven graphs constitute the cornerstone of many machine learning approaches. Recently, it was shown that sparse graphs (sparse representation based graphs) provide a powerful approach to graph-based semi-supervised classification. In this paper, we introduce a new structured sparse graph that is derived by integrating manifold-type constraints on the sparse coefficients without any a priori graph or similarity matrix. Furthermore, we introduce a direct and efficient solution to the proposed optimization problem. Unlike recent sparse graph construction methods that are based on the use of hand-crafted constraints or a predefined reference similarity matrix, our constraints are directly defined on the graph weights themselves, and can provide additional information to both local and global structures of the sparse graph. Experiments conducted on several image databases show that the proposed graph can give better results than many state-of-the-art sparse graphs when applied to the problem of graph-based label propagation. [ABSTRACT FROM AUTHOR]

Graph-based semi-supervised classification is widely used because it effectively exploits the characteristics of unlabeled data. However, the existing methods have a drawback in that they do not account for the inherent noise of the data. Noise in graph data refers to nodes that are isolated from classes and overlapping nodes between different classes. Therefore, neglecting noise can distort the data manifold, leading to over-smoothing and overall performance degradation. In this paper, we propose a noise-robust model called the dynamic shaving label propagation algorithm. Our proposed method comprises three parts: graph construction, noise definition and cutting, and label propagation. First, in the graph construction process, k is determined at the point when reverse nearest neighbors are identified for the most isolated nodes. Second, the noise definition process identifies the nodes classified as noise based on the number and distance of their reverse nearest neighbors. Finally, label propagation is performed dynamically and iteratively by adjusting the noise removal intensity. Using various simulated and real-world datasets, we evaluate the accuracy and noise robustness of the proposed method with those of existing methods to evaluate its effectiveness and applicability. The comparison results demonstrate that the proposed method outperforms the existing alternatives. • Label propagation is a representative graph-based semi-supervised method. • Reverse nearest neighbor information is used to build a k -nearest neighbors graph. • Considering noise can alleviate oversmoothing problem. • Properly defining and shaving noise improves the performance of noisy data. [ABSTRACT FROM AUTHOR]