Semi-supervised Learning Process Research Articles

Centroid Neural Network with Pairwise Constraints for Semi-supervised Learning

A clustering algorithm for datasets with pairwise constraints using the Centroid Neural Network (Cent.NN) is proposed in this paper. The proposed algorithm, referred to as the Centroid Neural Network with Pairwise Constraints (Cent. NN-PC) algorithm, utilizes Cent.NN as its backbone algorithm for data clustering and adopts a semi-supervised learning process for pairwise constraints. A newly formulated energy function is adopted from the original Cent.NN algorithm for the proposed Cent.NN-PC algorithm, introducing penalty terms for violating constraints. The weight update procedure of the proposed Cent.NN-PC algorithm finds optimal prototypes for the given dataset that minimize the quantization error while minimizing the number of violated constraints. In order to evaluate the performance of the proposed Cent.NN-PC algorithm, experiments on six different datasets from the UCI database and two bioinformatics datasets from the KEEL repository are carried out. The performance of the proposed algorithm is compared to that of the the Linear Constrained Vector Quantization Error (LCVQE) algorithm, one of the most commonly used algorithms for data clustering with pairwise constraints. In the experiments, five different numbers of pairwise constraints are utilized to evaluate the clustering performance with constraints of different sizes. The results show that the proposed Cent.NN-PC algorithm outperforms the LCVQE algorithm on most performance criteria, including the total quantization error, the number of violated constraints, and on the three performance metrics of the classification accuracy rate, F-score, and NMI measure outcome. The experiments also show that Cent.NN-PC provides much more stable clustering results at an improved operational speed compared to LCVQE.

LitPathExplorer: a confidence-based visual text analytics tool for exploring literature-enriched pathway models.

Pathway models are valuable resources that help us understand the various mechanisms underpinning complex biological processes. Their curation is typically carried out through manual inspection of published scientific literature to find information relevant to a model, which is a laborious and knowledge-intensive task. Furthermore, models curated manually cannot be easily updated and maintained with new evidence extracted from the literature without automated support. We have developed LitPathExplorer, a visual text analytics tool that integrates advanced text mining, semi-supervised learning and interactive visualization, to facilitate the exploration and analysis of pathway models using statements (i.e. events) extracted automatically from the literature and organized according to levels of confidence. LitPathExplorer supports pathway modellers and curators alike by: (i) extracting events from the literature that corroborate existing models with evidence; (ii) discovering new events which can update models; and (iii) providing a confidence value for each event that is automatically computed based on linguistic features and article metadata. Our evaluation of event extraction showed a precision of 89% and a recall of 71%. Evaluation of our confidence measure, when used for ranking sampled events, showed an average precision ranging between 61 and 73%, which can be improved to 95% when the user is involved in the semi-supervised learning process. Qualitative evaluation using pair analytics based on the feedback of three domain experts confirmed the utility of our tool within the context of pathway model exploration. LitPathExplorer is available at http://nactem.ac.uk/LitPathExplorer_BI/. sophia.ananiadou@manchester.ac.uk. Supplementary data are available at Bioinformatics online.

A multi-view semi-supervised approach for task-level web search success evaluation

Web search success evaluation is an effective way to evaluate the performance of search engines from the perspective of search experience. Many research efforts have been made to evaluate web search success via modeling user search behavior by analyzing search logs. Most of these studies consider the web search success evaluation as a binary classification problem, and use supervised learning approaches to evaluate whether a search experience is successful or not, which often require a large number of labeled data to learn accurately. Since unlabeled data of user search behaviors are easily obtainable in search logs, semi-supervised approaches have been exploited to improve the performance of web search success evaluation via combining labeled data and unlabeled data. However, the existing semi-supervised web search success evaluation approach would suffer from the model assumption violation, i.e., when the assumption of the model is not correct, training the web search success evaluation model with large amounts of unlabeled data would hurt the evaluation performance. In order to address this problem, a Multi-view Semi-Supervised web Search Success Evaluation (M4SE) approach is proposed, which exploits a multi-view mechanism during the semi-supervised learning process of web search success evaluation. M4SE considers the transitions between any two actions as the action view, and treats the dwell time between contiguous actions as the time view. M4SE uses the strategy of different parameter configurations on action and time views to generate diverse classifiers. Experiments on different search log datasets show that the proposed approach achieves better performance than the state-of-the-art semi-supervised approach for evaluating web search success.

Image Attribute Adaptation

Visual attributes can be considered as a middle-level semantic cue that bridges the gap between low-level image features and high-level object classes. Thus, attributes have the advantage of transcending specific semantic categories or describing objects across categories. Since attributes are often human-nameable and domain specific, much work constructs attribute annotations ad hoc or take them from an application-dependent ontology. To facilitate other applications with attributes, it is necessary to develop methods which can adapt a well-defined set of attributes to novel images. In this paper, we propose a framework for image attribute adaptation. The goal is to automatically adapt the knowledge of attributes from a well-defined auxiliary image set to a target image set, thus assisting in predicting appropriate attributes for target images. In the proposed framework, we use a non-linear mapping function corresponding to multiple base kernels to map each training images of both the auxiliary and the target sets to a Reproducing Kernel Hilbert Space (RKHS), where we reduce the mismatch of data distributions between auxiliary and target images. In order to make use of un-labeled images, we incorporate a semi-supervised learning process. We also introduce a robust loss function into our framework to remove the shared irrelevance and noise of training images. Experiments on two couples of auxiliary-target image sets demonstrate that the proposed framework has better performance of predicting attributes for target testing images, compared to three baselines and two state-of-the-art domain adaptation methods.

3D model retrieval and classification by semi-supervised learning with content-based similarity

The rapid development of 3D digital technology has led to an increasing volume of 3D model data. In addressing the management of such large scale data, effective content-based 3D model retrieval and recognition methods are highly desirable. In 3D model retrieval and recognition tasks, the distance measure between two 3D models plays an important role. In this paper, we propose a novel 3D model retrieval and recognition method that employs both a distance histogram and 3D moment invariants as features that are invariant to 3D object scaling, translation, and rotation. Disjoint information is used to measure the distance between the feature histograms, and the Euclidean distance is applied in calculating the distance between two moment features. These measures are then combined as the 3D model distance. Using this distance measure, the relationships between all 3D models in the dataset are formulated as a graph structure. A semi-supervised learning process is then conducted to estimate the relevance among the 3D models, and this is employed for 3D model retrieval and classification. To evaluate the effectiveness of the proposed method, we conduct experiments on two datasets. Experimental results and a comparison with state-of-the-art methods demonstrate that the proposed method achieves improved performance for 3D model retrieval and recognition tasks.

A semi-supervised incremental learning method based on adaptive probabilistic hypergraph for video semantic detection

Semantic categorization for the complex videos is an ambiguous task. The semi-supervised learning method based on hypergraph model can achieve multi-semantics labels, but a hypergraph model is sensitive to the radius parameter when it is constructed and the number of vertices belonging to a hyperedge is fixed. In this paper, a semi-supervised incremental learning method based on adaptive probabilistic hypergraph for video semantic detection is presented. In the probabilistic hypergraph modeling, a formula is presented as a measurement to adaptively decide whether a vertex is belonged to a hyperedge or not. The model has high robustness and can overcome the defect of fixed number of vertices belonging to the same hyperedge in the traditional probabilistic hypergraph model. In the semi-supervised incremental learning process, a threshold is defined, which is used to judge whether unlabeled sample can be added into the modeling, in order that the model learning result for unlabeled samples has high certainty. The experimental results show that our method can improve the model generalization ability and utilize the unlabeled samples effectively. In the aspects of recall rate and precision rate for semantic video concept detection from complex videos, our proposed method outperforms the compared methods.

Object Tracking Based on Online Classification Boosted by Discriminative Features

We propose a robust object tracking algorithm under “tracking by detection” framework. A sparse random Gaussian-distributed measurement matrix is used to build an object appearance model in the compressed domain, and detection is done by our semi-supervised learning process. Our method trains a group of weaker classifiers corresponding to every feature, and linearly combines them using the weights computed by its discriminative power to build a strong classification function, which detects a target from background. Kalman filtering is also adopted to smoothen the tracking results. Experiments show the proposed classification improves performance in tracking without drift.

Semi-Supervised Classification of Patient Safety Event Reports

The Veterans Health Administration patient safety reporting system receives more than 100,000 reports annually. The information contained in these reports is primarily in the form of natural language text. Improving the ability to efficiently mine these patient safety reports for information is the objective of a proposed semi-supervised method. A semi-supervised classification method leverages information from both labeled and unlabeled reports to predict categories for the unlabeled reports. Two different scenarios involving a semi-supervised learning process are examined, and both demonstrate good predictive results. The semi-supervised method shows much promise in assisting researchers and analysts toward accurately and more quickly separating reports of varying and often overlapping topics. The method is able to use the "stories" provided in patient safety reports to extend existing patient safety taxonomies beyond their static design.

A Boosted Co-Training Algorithm for Human Action Recognition

This paper proposes a boosted co-training algorithm for human action recognition. To address the view-sufficiency and view-dependency issues in co-training, two new confidence measures, namely, inter-view confidence and intra-view confidence, are proposed. They are dynamically fused into a semi-supervised learning process. Mutual information is employed to quantify the inter-view uncertainty and measure the independence among respective views. Intra-view confidence is estimated from boosted hypotheses to measure the total data inconsistency of labeled data and unlabeled data. Given a small set of labeled videos and a large set of unlabeled videos, the proposed semi-supervised learning algorithm trains a classifier by maximizing the inter-view confidence and intra-view confidence, and dynamically incorporating unlabeled data into the labeled data set. To evaluate the proposed boosted co-training algorithm, eigen-action and information saliency feature vectors are employed as two input views. The KTH and Weizmann human action databases are used for experiments, average recognition accuracy of 93.2% and 99.6% are obtained, respectively.

Semi-supervised learning by disagreement

In many real-world tasks, there are abundant unlabeled examples but the number of labeled training examples is limited, because labeling the examples requires human efforts and expertise. So, semi-supervised learning which tries to exploit unlabeled examples to improve learning performance has become a hot topic. Disagreement-based semi-supervised learning is an interesting paradigm, where multiple learners are trained for the task and the disagreements among the learners are exploited during the semi-supervised learning process. This survey article provides an introduction to research advances in this paradigm.

Hierarchical Wireless Multimedia Sensor Networks for Collaborative Hybrid Semi-Supervised Classifier Learning.

Wireless multimedia sensor networks (WMSN) have recently emerged as one ofthe most important technologies, driven by the powerful multimedia signal acquisition andprocessing abilities. Target classification is an important research issue addressed in WMSN,which has strict requirement in robustness, quickness and accuracy. This paper proposes acollaborative semi-supervised classifier learning algorithm to achieve durative onlinelearning for support vector machine (SVM) based robust target classification. The proposedalgorithm incrementally carries out the semi-supervised classifier learning process inhierarchical WMSN, with the collaboration of multiple sensor nodes in a hybrid computingparadigm. For decreasing the energy consumption and improving the performance, somemetrics are introduced to evaluate the effectiveness of the samples in specific sensor nodes,and a sensor node selection strategy is also proposed to reduce the impact of inevitablemissing detection and false detection. With the ant optimization routing, the learningprocess is implemented with the selected sensor nodes, which can decrease the energyconsumption. Experimental results demonstrate that the collaborative hybrid semi-supervised classifier learning algorithm can effectively implement target classification inhierarchical WMSN. It has outstanding performance in terms of energy efficiency and timecost, which verifies the effectiveness of the sensor nodes selection and ant optimizationrouting.

Software quality estimation with limited fault data: a semi-supervised learning perspective

We addresses the important problem of software quality analysis when there is limited software fault or fault-proneness data. A software quality model is typically trained using software measurement and fault data obtained from a previous release or similar project. Such an approach assumes that fault data is available for all the training modules. Various issues in software development may limit the availability of fault-proneness data for all the training modules. Consequently, the available labeled training dataset is such that the trained software quality model may not provide predictions. More specifically, the small set of modules with known fault-proneness labels is not sufficient for capturing the software quality trends of the project. We investigate semi-supervised learning with the Expectation Maximization (EM) algorithm for software quality estimation with limited fault-proneness data. The hypothesis is that knowledge stored in software attributes of the unlabeled program modules will aid in improving software quality estimation. Software data collected from a large NASA software project is used during the semi-supervised learning process. The software quality model is evaluated with multiple test datasets collected from other NASA software projects. Compared to software quality models trained only with the available set of labeled program modules, the EM-based semi-supervised learning scheme improves generalization performance of the software quality models.