Weak-label Learning Research Articles

Multi-label weak-label learning via semantic reconstruction and label correlations

In the multi-label classification task, an instance is simultaneously associated with multiple semantic labels. Due to the high complexity of the semantic space in practical applications, obtaining instances with full labels is difficult, leading to the weak-label problem. Existing methods focus on the low-rank and instance manifold regularization assumptions of the label matrix to recover the ground-truth label matrix but ignore the influence that the above assumptions may not hold due to the semantic noise caused by missing labels. To address the problem, this paper proposes a simple and effective method to recover the label space by reconstructing the label semantic space through joint label correlation to solve the multi-label weak-label classification task. Specifically, we leverage the label information consistency and feature-label dependency assumptions to reconstruct the semantic space and consider label correlations to enhance the information of semantic views. Moreover, l2,1-norm is utilized to mitigate the effect of missing label space noise. Additionally, the linear model of the proposed method is expanded to a nonlinear version of the kernel method to address the problem of the inseparability of linear data. Extensive experiments on several real-world tasks show that the proposed method outperforms some state-of-the-art methods.

DeepSMILE: Contrastive self-supervised pre-training benefits MSI and HRD classification directly from H&E whole-slide images in colorectal and breast cancer.

We propose a Deep learning-based weak label learning method for analyzing whole slide images (WSIs) of Hematoxylin and Eosin (H&E) stained tumor tissue not requiring pixel-level or tile-level annotations using Self-supervised pre-training and heterogeneity-aware deep Multiple Instance LEarning (DeepSMILE). We apply DeepSMILE to the task of Homologous recombination deficiency (HRD) and microsatellite instability (MSI) prediction. We utilize contrastive self-supervised learning to pre-train a feature extractor on histopathology tiles of cancer tissue. Additionally, we use variability-aware deep multiple instance learning to learn the tile feature aggregation function while modeling tumor heterogeneity. For MSI prediction in a tumor-annotated and color normalized subset of TCGA-CRC (n=360 patients), contrastive self-supervised learning improves the tile supervision baseline from 0.77 to 0.87 AUROC, on par with our proposed DeepSMILE method. On TCGA-BC (n=1041 patients) without any manual annotations, DeepSMILE improves HRD classification performance from 0.77 to 0.81 AUROC compared to tile supervision with either a self-supervised or ImageNet pre-trained feature extractor. Our proposed methods reach the baseline performance using only 40% of the labeled data on both datasets. These improvements suggest we can use standard self-supervised learning techniques combined with multiple instance learning in the histopathology domain to improve genomic label classification performance with fewer labeled data.

Learning from Weak-Label Data: A Deep Forest Expedition

Weak-label learning deals with the problem where each training example is associated with multiple ground-truth labels simultaneously but only partially provided. This circumstance is frequently encountered when the number of classes is very large or when there exists a large ambiguity between class labels, and significantly influences the performance of multi-label learning. In this paper, we propose LCForest, which is the first tree ensemble based deep learning method for weak-label learning. Rather than formulating the problem as a regularized framework, we employ the recently proposed cascade forest structure, which processes information layer-by-layer, and endow it with the ability of exploiting from weak-label data by a concise and highly efficient label complement structure. Specifically, in each layer, the label vector of each instance from testing-fold is modified with the predictions of random forests trained with the corresponding training-fold. Since the ground-truth label matrix is inaccessible, we can not estimate the performance via cross-validation directly. In order to control the growth of cascade forest, we adopt label frequency estimation and the complement flag mechanism. Experiments show that the proposed LCForest method compares favorably against the existing state-of-the-art multi-label and weak-label learning methods.

Learning Multi Labels from Single Label —— An Extreme Weak Label Learning Algorithm

This paper presents a novel algorithm for an extreme form of weak label learning, in which only one of all relevant labels is given for each training sample. Using genetic algorithm, all of the labels in the training set are optimally divided into several non-overlapping groups to maximize the label distinguishability in every group. Multiple classifiers are trained separately and ensembled for label predictions. Experimental results show significant improvement over previous weak label learning algorithms.

Learning From Semi-Supervised Weak-Label Data

Multi-label learning deals with data objects associated with multiple labels simultaneously. Previous studies typically assume that for each instance, the full set of relevant labels associated with each training instance is given. In many applicationssuch as image annotation, however, it’s usually difficult to get the full label set for each instance and only a partial or even empty set of relevant labels is available. We call this kind of problem as "semi-supervised weak-label learning" problem. In this work we propose the SSWL (Semi-Supervised Weak-Label) method to address this problem. Both instance similarity and label similarity are considered for the complement of missing labels. Ensemble of multiple models are utilized to improve the robustness when label information is insufficient. We formulate the objective as a bi-convex optimization problem with an efficient block coordinate descent algorithm. Experiments validate the effectiveness of SSWL.

Expede Herculem: Learning Multi Labels From Single Label

Although there has been a lot of research in multi-label learning task, little attention has been paid on the weak label problem, in which only a subset of labels has been assigned to each instance in the training set. The extreme form of weak label learning is to predict all the labels from just one label set in the training phase. In this paper, we focus on dealing with this kind of weak label learning task, which is commonly met in old legacy information system, and it is also called “Hercules Learning.” We propose a label-group-optimization-based Hercules learning algorithm, which divides the entire label set into multiple groups according to the classifier’s capability to distinguish them, so for each group, we can train a classifier which can predict instance’s label within the group with high accuracy. The experimental results show that our algorithm is obviously superior to the existing weak label learning algorithm.

Learning safe multi-label prediction for weakly labeled data

In this paper we study multi-label learning with weakly labeled data, i.e., labels of training examples are incomplete, which commonly occurs in real applications, e.g., image classification, document categorization. This setting includes, e.g., (i) semi-supervised multi-label learning where completely labeled examples are partially known; (ii) weak label learning where relevant labels of examples are partially known; (iii) extended weak label learning where relevant and irrelevant labels of examples are partially known. Previous studies often expect that the learning method with the use of weakly labeled data will improve the performance, as more data are employed. This, however, is not always the cases in reality, i.e., weakly labeled data may sometimes degenerate the learning performance. It is desirable to learn safe multi-label prediction that will not hurt performance when weakly labeled data is involved in the learning procedure. In this work we optimize multi-label evaluation metrics ( $$\hbox {F}_1$$ score and Top-k precision) given that the ground-truth label assignment is realized by a convex combination of base multi-label learners. To cope with the infinite number of possible ground-truth label assignments, cutting-plane strategy is adopted to iteratively generate the most helpful label assignments. The whole optimization is cast as a series of simple linear programs in an efficient manner. Extensive experiments on three weakly labeled learning tasks, namely, (i) semi-supervised multi-label learning; (ii) weak label learning and (iii) extended weak label learning, clearly show that our proposal improves the safeness of using weakly labeled data compared with many state-of-the-art methods.

Semi-supervised multi-label classification using incomplete label information

Classifying multi-label instances using incompletely labeled instances is one of the fundamental tasks in multi-label learning. Most existing methods regard this task as supervised weak-label learning problem and assume sufficient partially labeled instances are available. However, collecting or annotating such instances is expensive and time-consuming. In contrast, abundant unlabeled instances are easy to accumulate. Recently, some methods move toward exploiting unlabeled instances and performing transductive multi-label classification. However, these methods can not directly apply to new instances, which are not available during training process. In this paper, we proposed an approach called Semi-supervised multi-label classification using incomplete label information (SMILE for short). SMILE first estimates label correlation from partially labeled instances and replenishes missing labels of these instances. Then, it takes advantage of labeled and unlabeled instances to construct a neighborhood graph. Next, the known labels and replenished ones of labeled instances, along with unlabeled instances are exploited to train a graph based semi-supervised linear classifier. SMILE can further replenish the missing labels of training instances based on the adopted neighborhood graph. In addition, it can directly predict the labels of completely unlabeled new instances. The empirical study on multi-label datasets shows that SMILE performs significantly better than other related methods across various evaluation criteria and it is important to leverage unlabeled data with label correlation for multi-label classification.

Multi-instance multilabel learning with weak-label for predicting protein function in electricigens.

Nature often brings several domains together to form multidomain and multifunctional proteins with a vast number of possibilities. In our previous study, we disclosed that the protein function prediction problem is naturally and inherently Multi-Instance Multilabel (MIML) learning tasks. Automated protein function prediction is typically implemented under the assumption that the functions of labeled proteins are complete; that is, there are no missing labels. In contrast, in practice just a subset of the functions of a protein are known, and whether this protein has other functions is unknown. It is evident that protein function prediction tasks suffer from weak-label problem; thus protein function prediction with incomplete annotation matches well with the MIML with weak-label learning framework. In this paper, we have applied the state-of-the-art MIML with weak-label learning algorithm MIMLwel for predicting protein functions in two typical real-world electricigens organisms which have been widely used in microbial fuel cells (MFCs) researches. Our experimental results validate the effectiveness of MIMLwel algorithm in predicting protein functions with incomplete annotation.

Protein Function Prediction with Incomplete Annotations.

Automated protein function prediction is one of the grand challenges in computational biology. Multi-label learning is widely used to predict functions of proteins. Most of multi-label learning methods make prediction for unlabeled proteins under the assumption that the labeled proteins are completely annotated, i.e., without any missing functions. However, in practice, we may have a subset of the ground-truth functions for a protein, and whether the protein has other functions is unknown. To predict protein functions with incomplete annotations, we propose a Protein Function Prediction method with Weak-label Learning (ProWL) and its variant ProWL-IF. Both ProWL and ProWL-IF can replenish the missing functions of proteins. In addition, ProWL-IF makes use of the knowledge that a protein cannot have certain functions, which can further boost the performance of protein function prediction. Our experimental results on protein-protein interaction networks and gene expression benchmarks validate the effectiveness of both ProWL and ProWL-IF.

Multi-Label Learning with Weak Label

Multi-label learning deals with data associated with multiple labels simultaneously. Previous work on multi-label learning assumes that for each instance, the “full” label set associated with each training instance is given by users. In many applications, however, to get the full label set for each instance is difficult and only a “partial” set of labels is available. In such cases, the appearance of a label means that the instance is associated with this label, while the absence of a label does not imply that this label is not proper for the instance. We call this kind of problem “weak label” problem. In this paper, we propose the WELL (WEak Label Learning) method to solve the weak label problem. We consider that the classification boundary for each label should go across low density regions, and that each label generally has much smaller number of positive examples than negative examples. The objective is formulated as a convex optimization problem which can be solved efficiently. Moreover, we exploit the correlation between labels by assuming that there is a group of low-rank base similarities, and the appropriate similarities between instances for different labels can be derived from these base similarities. Experiments validate the performance of WELL.