To date, many machine learning applications have multiple views of features, and different applications require specific multivariate performance measures, such as the F-score for retrieval. However, existing multivariate performance measure optimization methods are limited to single-view data, while traditional multi-view learning methods cannot optimize multivariate performance measures directly. To fill this gap, in this paper, we propose the problem of optimizing multivariate performance measures from multi-view data, and an effective method to solve it. We propose to learn linear discriminant functions for different views, and combine them to construct an overall multivariate mapping function for multi-view data. To learn the parameters of the linear discriminant functions of different views to optimize a given multivariate performance measure, we formulate an optimization problem. In this problem, we propose to minimize the complexity of the linear discriminant function of each view, promote the consistency of the responses of different views over the same data points, and minimize the upper boundary of the corresponding loss of a given multivariate performance measure. To optimize this problem, we develop an iterative cutting-plane algorithm. Experiments on four benchmark data sets show that it not only outperforms traditional single-view based multivariate performance optimization methods, but also achieves better results than ordinary multi-view learning methods.

In this paper we study the problem of content-based image retrieval. In this problem, the most popular performance measure is the top precision measure, and the most important component of a retrieval system is the similarity function used to compare a query image against a database image. However, up to now, there is no existing similarity learning method proposed to optimize the top precision measure. To fill this gap, in this paper, we propose a novel similarity learning method to maximize the top precision measure. We model this problem as a minimization problem with an objective function as the combination of the losses of the relevant images ranked behind the top-ranked irrelevant image, and the squared Frobenius norm of the similarity function parameter. This minimization problem is solved as a quadratic programming problem. The experiments over two benchmark data sets show the advantages of the proposed method over other similarity learning methods when the top precision is used as the performance measure.

In this paper, we propose a novel learning framework for the problem of domain transfer learning. We map the data of two domains to one single common space, and learn a classifier in this common space. Then we adapt the common classifier to the two domains by adding two adaptive functions to it respectively. In the common space, the target domain data points are weighted and matched to the target domain in term of distributions. The weighting terms of source domain data points and the target domain classification responses are also regularized by the local reconstruction coefficients. The novel transfer learning framework is evaluated over some benchmark cross-domain data sets, and it outperforms the existing state-of-the-art transfer learning methods.

MOTIVATION: Protein homology detection, a fundamental problem in computational biology, is an indispensable step toward predicting protein structures and understanding protein functions. Despite the advances in recent decades on sequence alignment, threading and alignment-free methods, protein homology detection remains a challenging open problem. Recently, network methods that try to find transitive paths in the protein structure space demonstrate the importance of incorporating network information of the structure space. Yet, current methods merge the sequence space and the structure space into a single space, and thus introduce inconsistency in combining different sources of information. METHOD: We present a novel network-based protein homology detection method, CMsearch, based on cross-modal learning. Instead of exploring a single network built from the mixture of sequence and structure space information, CMsearch builds two separate networks to represent the sequence space and the structure space. It then learns sequence-structure correlation by simultaneously taking sequence information, structure information, sequence space information and structure space information into consideration. RESULTS: We tested CMsearch on two challenging tasks, protein homology detection and protein structure prediction, by querying all 8332 PDB40 proteins. Our results demonstrate that CMsearch is insensitive to the similarity metrics used to define the sequence and the structure spaces. By using HMM-HMM alignment as the sequence similarity metric, CMsearch clearly outperforms state-of-the-art homology detection methods and the CASP-winning template-based protein structure prediction methods. AVAILABILITY AND IMPLEMENTATION: Our program is freely available for download from http://sfb.kaust.edu.sa/Pages/Software.aspx CONTACT: : [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Multivariate performance measure optimization refers to learning predictive models such that a desired complex performance measure can be optimized over a training set, such as the F1 score. Up to now, all the existing multivariate performance measure optimization methods are limited to a completely labeled data tuple, i.e., the label tuple is complete. However, in real-world applications, sometimes it is difficult to obtain a complete label tuple. In this paper, we show that the multivariate performance measures can also be optimized by learning from partially labeled data tuple, when the label tuple is incomplete. We introduce a slack label tuple to represent the sought complete true label tuple, and learn it jointly with a hyper predictor, so that it can be consistent to the known labels, prediction results, and is smooth in the neighborhood. We develop an iterative learning algorithm to learn the slack label tuple and the hyper predictor. Its advantage over state-of-the-art multivariate performance measure optimization methods is shown by experiments on benchmark data sets.

In this paper we study the problem of learning from multiple modal data for purpose of document classification. In this problem, each document is composed two different modals of data, i.e., An image and a text. Cross-modal factor analysis (CFA) has been proposed to project the two different modals of data to a shared data space, so that the classification of a image or a text can be performed directly in this space. A disadvantage of CFA is that it has ignored the supervision information. In this paper, we improve CFA by incorporating the supervision information to represent and classify both image and text modals of documents. We project both image and text data to a shared data space by factor analysis, and then train a class label predictor in the shared space to use the class label information. The factor analysis parameter and the predictor parameter are learned jointly by solving one single objective function. With this objective function, we minimize the distance between the projections of image and text of the same document, and the classification error of the projection measured by hinge loss function. The objective function is optimized by an alternate optimization strategy in an iterative algorithm. Experiments in two different multiple modal document data sets show the advantage of the proposed algorithm over other CFA methods.

Nonnegative matrix factorization (NMF), a popular part-based representation technique, does not capture the intrinsic local geometric structure of the data space. Graph regularized NMF (GNMF) was recently proposed to avoid this limitation by regularizing NMF with a nearest neighbor graph constructed from the input data set. However, GNMF has two main bottlenecks. First, using the original feature space directly to construct the graph is not necessarily optimal because of the noisy and irrelevant features and nonlinear distributions of data samples. Second, one possible way to handle the nonlinear distribution of data samples is by kernel embedding. However, it is often difficult to choose the most suitable kernel. To solve these bottlenecks, we propose two novel graph-regularized NMF methods, AGNMFFS and AGNMFMK, by introducing feature selection and multiple-kernel learning to the graph regularized NMF, respectively. Instead of using a fixed graph as in GNMF, the two proposed methods learn the nearest neighbor graph that is adaptive to the selected features and learned multiple kernels, respectively. For each method, we propose a unified objective function to conduct feature selection/multi-kernel learning, NMF and adaptive graph regularization simultaneously. We further develop two iterative algorithms to solve the two optimization problems. Experimental results on two challenging pattern classification tasks demonstrate that the proposed methods significantly outperform state-of-the-art data representation methods.

Nonnegative Matrix Factorization (NMF) has been a popular representation method for pattern classification problems. It tries to decompose a nonnegative matrix of data samples as the product of a nonnegative basis matrix and a nonnegative coefficient matrix. The columns of the coefficient matrix can be used as new representations of these data samples. However, traditional NMF methods ignore class labels of the data samples. In this paper, we propose a novel supervised NMF algorithm to improve the discriminative ability of the new representation by using the class labels. Using the class labels, we separate all the data sample pairs into within-class pairs and between-class pairs. To improve the discriminative ability of the new NMF representations, we propose to minimize the maximum distance of the within-class pairs in the new NMF space, and meanwhile to maximize the minimum distance of the between-class pairs. With this criterion, we construct an objective function and optimize it with regard to basis and coefficient matrices, and slack variables alternatively, resulting in an iterative algorithm. The proposed algorithm is evaluated on three pattern classification problems and experiment results show that it outperforms the state-of-the-art supervised NMF methods.

A combination of the sparse coding and transfer learning techniques was shown to be accurate and robust in classification tasks where training and testing objects have a shared feature space but are sampled from different underlying distributions, i.e., belong to different domains. The key assumption in such case is that in spite of the domain disparity, samples from different domains share some common hidden factors. Previous methods often assumed that all the objects in the target domain are unlabeled, and thus the training set solely comprised objects from the source domain. However, in real world applications, the target domain often has some labeled objects, or one can always manually label a small number of them. In this paper, we explore such possibility and show how a small number of labeled data in the target domain can significantly leverage classification accuracy of the state-of-the-art transfer sparse coding methods. We further propose a unified framework named supervised transfer sparse coding (STSC) which simultaneously optimizes sparse representation, domain transfer and classification. Experimental results on three applications demonstrate that a little manual labeling and then learning the model in a supervised fashion can significantly improve classification accuracy.

In this paper, we propose a novel image set representation and classification method by maximizing the margin of image sets. The margin of an image set is defined as the difference of the distance to its nearest image set from different classes and the distance to its nearest image set of the same class. By modeling the image sets by using both their image samples and their affine hull models, and maximizing the margins of the images sets, the image set representation parameter learning problem is formulated as an minimization problem, which is further optimized by an expectation - maximization (EM) strategy with accelerated proximal gradient (APG) optimization in an iterative algorithm. To classify a given test image set, we assign it to the class which could provide the largest margin. Experiments on two applications of video-sequence-based face recognition demonstrate that the proposed method significantly outperforms state-of-the-art image set classification methods in terms of both effectiveness and efficiency.

Sparse coding approximates the data sample as a sparse linear combination of some basic codewords and uses the sparse codes as new presentations. In this paper, we investigate learning discriminative sparse codes by sparse coding in a semi-supervised manner, where only a few training samples are labeled. By using the manifold structure spanned by the data set of both labeled and unlabeled samples and the constraints provided by the labels of the labeled samples, we learn the variable class labels for all the samples. Furthermore, to improve the discriminative ability of the learned sparse codes, we assume that the class labels could be predicted from the sparse codes directly using a linear classifier. By solving the codebook, sparse codes, class labels and classifier parameters simultaneously in a unified objective function, we develop a semi-supervised sparse coding algorithm. Experiments on two real-world pattern recognition problems demonstrate the advantage of the proposed methods over supervised sparse coding methods on partially labeled data sets.

Domain transfer learning aims to learn an effective classifier for a target domain, where only a few labeled samples are available, with the help of many labeled samples from a source domain. The source and target domain samples usually share the same features and class label space, but have significantly different In these experiments error of the classifier distributions. Nonnegative Matrix Factorization (NMF) has been studied and applied widely as a powerful data representation method. However, NMF is limited to single domain learning problem. It can not be directly used in domain transfer learning problem due to the significant differences between the distributions of the source and target domains. In this paper, we extend the NMF method to domain transfer learning problem. The Maximum Mean Discrepancy (MMD) criteria is employed to reduce the mismatch of source and target domain distributions in the coding vector space. Moreover, we also learn a classifier in the coding vector space to directly utilize the class labels from both the two domains. We construct an unified objective function for the learning of both NMF parameters and classifier parameters, which is optimized alternately in an iterative algorithm. The proposed algorithm is evaluated on two challenging domain transfer tasks, and the encouraging experimental results show its advantage over state-of-the-art domain transfer learning algorithms.

Local-learning-based feature selection has been successfully applied to high-dimensional data analysis. It utilizes class labels to define a margin for each data sample and selects the most discriminative features by maximizing the margins with regard to a feature weight vector. However, it requires that all data samples are labeled, which makes it unsuitable for semi-supervised learning where only a handful of training samples are labeled while most are unlabeled. To address this issue, we herein propose a new semi-supervised local-learning-based feature selection method. The basic idea is to learn the class labels of unlabeled samples in a new feature subspace induced by the learned feature weights, and then use the learned class labels to define the margins for feature weight learning. By constructing and optimizing a unified objective function, the feature weights and class labels are learned simultaneously in an iterative algorithm. The experiments performed on some benchmark data sets show the advantage of the proposed algorithm over stat-of-the-art semi-supervised feature selection methods.

Sparse coding has shown its power as an effective data representation method. However, up to now, all the sparse coding approaches are limited within the single domain learning problem. In this paper, we extend the sparse coding to cross domain learning problem, which tries to learn from a source domain to a target domain with significant different distribution. We impose the Maximum Mean Discrepancy (MMD) criterion to reduce the cross-domain distribution difference of sparse codes, and also regularize the sparse codes by the class labels of the samples from both domains to increase the discriminative ability. The encouraging experiment results of the proposed cross-domain sparse coding algorithm on two challenging tasks --- image classification of photograph and oil painting domains, and multiple user spam detection --- show the advantage of the proposed method over other cross-domain data representation methods.

BACKGROUND: Non-negative matrix factorization (NMF) has been shown to be a powerful tool for clustering gene expression data, which are widely used to classify cancers. NMF aims to find two non-negative matrices whose product closely approximates the original matrix. Traditional NMF methods minimize either the l2 norm or the Kullback-Leibler distance between the product of the two matrices and the original matrix. Correntropy was recently shown to be an effective similarity measurement due to its stability to outliers or noise. RESULTS: We propose a maximum correntropy criterion (MCC)-based NMF method (NMF-MCC) for gene expression data-based cancer clustering. Instead of minimizing the l2 norm or the Kullback-Leibler distance, NMF-MCC maximizes the correntropy between the product of the two matrices and the original matrix. The optimization problem can be solved by an expectation conditional maximization algorithm. CONCLUSIONS: Extensive experiments on six cancer benchmark sets demonstrate that the proposed method is significantly more accurate than the state-of-the-art methods in cancer clustering.

Sparse coding has been popularly used as an effective data representation method in various applications, such as computer vision, medical imaging and bioinformatics. However, the conventional sparse coding algorithms and their manifold-regularized variants (graph sparse coding and Laplacian sparse coding), learn codebooks and codes in an unsupervised manner and neglect class information that is available in the training set. To address this problem, we propose a novel discriminative sparse coding method based on multi-manifolds, that learns discriminative class-conditioned codebooks and sparse codes from both data feature spaces and class labels. First, the entire training set is partitioned into multiple manifolds according to the class labels. Then, we formulate the sparse coding as a manifold–manifold matching problem and learn class-conditioned codebooks and codes to maximize the manifold margins of different classes. Lastly, we present a data sample-manifold matching-based strategy to classify the unlabeled data samples. Experimental results on somatic mutations identification and breast tumor classification based on ultrasonic images demonstrate the efficacy of the proposed data representation and classification approach.