In this work we propose a method for reducing the dimensionality of tensor objects in a binary classification framework. The proposed Common Mode Patterns method takes into consideration the labels' information, and ensures that tensor objects that belong to different classes do not share common features after the reduction of their dimensionality. We experimentally validate the proposed supervised subspace learning technique and compared it against Multilinear Principal Component Analysis using a publicly available hyperspectral imaging dataset. Experimental results indicate that the proposed CMP method can efficiently reduce the dimensionality of tensor objects, while, at the same time, increasing the inter-class separability.
Insufficient training data and severe class imbalance are often limiting factors when developing machine learning models for the classification of rare diseases. In this work, we address the problem of classifying bone lesions from X-ray images by increasing the small number of positive samples in the training set. We propose a generative data augmentation approach based on a cycle-consistent generative adversarial network that synthesizes bone lesions on images without pathology. We pose the generative task as an image-patch translation problem that we optimize specifically for distinct bones (humerus, tibia, femur). In experimental results, we confirm that the described method mitigates the class imbalance problem in the binary classification task of bone lesion detection. We show that the augmented training sets enable the training of superior classifiers achieving better performance on a held-out test set. Additionally, we demonstrate the feasibility of transfer learning and apply a generative model that was trained on one body part to another.
In classical machine learning, a set of weak classifiers can be adaptively combined to form a strong classifier for improving the overall performance, a technique called adaptive boosting (or AdaBoost). However, constructing the strong classifier for a large data set is typically resource consuming. Here we propose a quantum extension of AdaBoost, demonstrating a quantum algorithm that can output the optimal strong classifier with a quadratic speedup in the number of queries of the weak classifiers. Our results also include a generalization of the standard AdaBoost to the cases where the output of each classifier may be probabilistic even for the same input. We prove that the update rules and the query complexity of the non-deterministic classifiers are the same as those of deterministic classifiers, which may be of independent interest to the classical machine-learning community. Furthermore, the AdaBoost algorithm can also be applied to data encoded in the form of quantum states; we show how the training set can be simplified by using the tools of t-design. Our approach describes a model of quantum machine learning where quantum speedup is achieved in finding the optimal classifier, which can then be applied for classical machine-learning applications.
Motivation: Ab initio protein docking represents a major challenge for optimizing a noisy and costly "black box"-like function in a high-dimensional space. Despite progress in this field, there is no docking method available for rigorous uncertainty quantification (UQ) of its solution quality (e.g. interface RMSD or iRMSD). Results: We introduce a novel algorithm, Bayesian Active Learning (BAL), for optimization and UQ of such black-box functions and flexible protein docking. BAL directly models the posterior distribution of the global optimum (or native structures for protein docking) with active sampling and posterior estimation iteratively feeding each other. Furthermore, we use complex normal modes to represent a homogeneous Euclidean conformation space suitable for high-dimension optimization and construct funnel-like energy models for encounter complexes. Over a protein docking benchmark set and a CAPRI set including homology docking, we establish that BAL significantly improve against both starting points by rigid docking and refinements by particle swarm optimization, providing for one third targets a top-3 near-native prediction. BAL also generates tight confidence intervals with half range around 25% of iRMSD and confidence level at 85%. Its estimated probability of a prediction being native or not achieves binary classification AUROC at 0.93 and AUPRC over 0.60 (compared to 0.14 by chance); and also found to help ranking predictions. To the best of our knowledge, this study represents the first uncertainty quantification solution for protein docking, with theoretical rigor and comprehensive assessment. Source codes are available at https://github.com/Shen-Lab/BAL.
Gaussian conditional random fields (GCRF) are a well-known used structured model for continuous outputs that uses multiple unstructured predictors to form its features and at the same time exploits dependence structure among outputs, which is provided by a similarity measure. In this paper, a Gaussian conditional random fields model for structured binary classification (GCRFBC) is proposed. The model is applicable to classification problems with undirected graphs, intractable for standard classification CRFs. The model representation of GCRFBC is extended by latent variables which yield some appealing properties. Thanks to the GCRF latent structure, the model becomes tractable, efficient and open to improvements previously applied to GCRF regression models. In addition, the model allows for reduction of noise, that might appear if structures were defined directly between discrete outputs. Additionally, two different forms of the algorithm are presented: GCRFBCb (GCRGBC - Bayesian) and GCRFBCnb (GCRFBC - non Bayesian). The extended method of local variational approximation of sigmoid function is used for solving empirical Bayes in Bayesian GCRFBCb variant, whereas MAP value of latent variables is the basis for learning and inference in the GCRFBCnb variant. The inference in GCRFBCb is solved by Newton-Cotes formulas for one-dimensional integration. Both models are evaluated on synthetic data and real-world data. It was shown that both models achieve better prediction performance than unstructured predictors. Furthermore, computational and memory complexity is evaluated. Advantages and disadvantages of the proposed GCRFBCb and GCRFBCnb are discussed in detail.
Appropriately evaluating the discrepancy between domains is essential for the success of unsupervised domain adaptation. In this paper, we first point out that existing discrepancy measures are less informative when complex models such as deep neural networks are used, in addition to the facts that they can be computationally highly demanding and their range of applications is limited only to binary classification. We then propose a novel domain discrepancy measure, called the paired hypotheses discrepancy (PHD), to overcome these shortcomings. PHD is computationally efficient and applicable to multi-class classification. Through generalization error bound analysis, we theoretically show that PHD is effective even for complex models. Finally, we demonstrate the practical usefulness of PHD through experiments.
We present two sampled quasi-Newton methods (sampled LBFGS and sampled LSR1) for solving empirical risk minimization problems that arise in machine learning. Contrary to the classical variants of these methods that sequentially build Hessian or inverse Hessian approximations as the optimization progresses, our proposed methods sample points randomly around the current iterate at every iteration to produce these approximations. As a result, the approximations constructed make use of more reliable (recent and local) information, and do not depend on past iterate information that could be significantly stale. Our proposed algorithms are efficient in terms of accessed data points (epochs) and have enough concurrency to take advantage of parallel/distributed computing environments. We provide convergence guarantees for our proposed methods. Numerical tests on a toy classification problem as well as on popular benchmarking binary classification and neural network training tasks reveal that the methods outperform their classical variants.
Modern computing and communication technologies can make data collection procedures very efficient. However, our ability to analyze large data sets and/or to extract information out from them is hard-pressed to keep up with our capacities for data collection. Among these huge data sets, some of them are not collected for any particular research purpose. For a classification problem, this means that the essential label information may not be readily obtainable, in the data set in hands, and an extra labeling procedure is required such that we can have enough label information to be used for constructing a classification model. When the size of a data set is huge, to label each subject in it will cost a lot in both capital and time. Thus, it is an important issue to decide which subjects should be labeled first in order to efficiently reduce the training cost/time. Active learning method is a promising outlet for this situation, because with the active learning ideas, we can select the unlabeled subjects sequentially without knowing their label information. In addition, there will be no confirmed information about the essential variables for constructing an efficient classification rule. Thus, how to merge a variable selection scheme with an active learning procedure is of interest. In this paper, we propose a procedure for building binary classification models when the complete label information is not available in the beginning of the training stage. We study an model-based active learning procedure with sequential variable selection schemes, and discuss the results of the proposed procedure from both theoretical and numerical aspects.
We propose a general framework to learn deep generative models via \textbf{V}ariational \textbf{Gr}adient Fl\textbf{ow} (VGrow) on probability spaces. The evolving distribution that asymptotically converges to the target distribution is governed by a vector field, which is the negative gradient of the first variation of the $f$-divergence between them. We prove that the evolving distribution coincides with the pushforward distribution through the infinitesimal time composition of residual maps that are perturbations of the identity map along the vector field. The vector field depends on the density ratio of the pushforward distribution and the target distribution, which can be consistently learned from a binary classification problem. Connections of our proposed VGrow method with other popular methods, such as VAE, GAN and flow-based methods, have been established in this framework, gaining new insights of deep generative learning. We also evaluated several commonly used divergences, including Kullback-Leibler, Jensen-Shannon, Jeffrey divergences as well as our newly discovered `logD' divergence which serves as the objective function of the logD-trick GAN. Experimental results on benchmark datasets demonstrate that VGrow can generate high-fidelity images in a stable and efficient manner, achieving competitive performance with state-of-the-art GANs.
State-of-the-art neural networks are vulnerable to adversarial examples; they can easily misclassify inputs that are imperceptibly different than their training and test data. In this work, we establish that the use of cross-entropy loss function and the low-rank features of the training data have responsibility for the existence of these inputs. Based on this observation, we suggest that addressing adversarial examples requires rethinking the use of cross-entropy loss function and looking for an alternative that is more suited for minimization with low-rank features. In this direction, we present a training scheme called differential training, which uses a loss function defined on the differences between the features of points from opposite classes. We show that differential training can ensure a large margin between the decision boundary of the neural network and the points in the training dataset. This larger margin increases the amount of perturbation needed to flip the prediction of the classifier and makes it harder to find an adversarial example with small perturbations. We test differential training on a binary classification task with CIFAR-10 dataset and demonstrate that it radically reduces the ratio of images for which an adversarial example could be found -- not only in the training dataset, but in the test dataset as well.
In many real-world prediction tasks, class labels include information about the relative ordering between labels, which is not captured by commonly-used loss functions such as multi-category cross-entropy. Recently, the deep learning community adopted ordinal regression frameworks to take such ordering information into account. Neural networks were equipped with ordinal regression capabilities by transforming ordinal targets into binary classification subtasks. However, this method suffers from inconsistencies among the different binary classifiers. To resolve these inconsistencies, we propose the COnsistent RAnk Logits (CORAL) framework with strong theoretical guarantees for rank-monotonicity and consistent confidence scores. Moreover, the proposed method is architecture-agnostic and can extend arbitrary state-of-the-art deep neural network classifiers for ordinal regression tasks. The empirical evaluation of the proposed rank-consistent method on a range of face-image datasets for age prediction shows a substantial reduction of the prediction error compared to the reference ordinal regression network.
Existing method generates class activation map (CAM) by a set of fixed classes (i.e., using all the classes), while the discriminative cues between class pairs are not considered. Note that activation maps by considering different class pair are complementary, and therefore can provide more discriminative cues to overcome the shortcoming of the existing CAM generation that the highlighted regions are usually local part regions rather than global object regions due to the lack of object cues. In this paper, we generate CAM by using a few of representative classes, with aim of extracting more discriminative cues by considering each class pair to obtain CAM more globally. The advantages are twofold. Firstly, the representative classes are able to obtain activation regions that are complementary to each other, and therefore leads to generating activation map more accurately. Secondly, we only need to consider a small number of representative classes, making the CAM generation suitable for small networks. We propose a clustering based method to select the representative classes. Multiple binary classification models rather than a multiple class classification model are used to generate the CAM. Moreover, we propose a multi-layer fusion based CAM generation method to simultaneously combine high-level semantic features and low-level detail features. We validate the proposed method on the PASCAL VOC and COCO database in terms of segmentation groundtruth. Various networks such as classical network (Resnet-50, Resent-101 and Resnet-152) and small network (VGG-19, Resnet-18 and Mobilenet) are considered. Experimental results show that the proposed method improves the CAM generation obviously.
Playlist recommendation involves producing a set of songs that a user might enjoy. We investigate this problem in three cold-start scenarios: (i) cold playlists, where we recommend songs to form new personalised playlists for an existing user; (ii) cold users, where we recommend songs to form new playlists for a new user; and (iii) cold songs, where we recommend newly released songs to extend users' existing playlists. We propose a flexible multitask learning method to deal with all three settings. The method learns from user-curated playlists, and encourages songs in a playlist to be ranked higher than those that are not by minimising a bipartite ranking loss. Inspired by an equivalence between bipartite ranking and binary classification, we show how one can efficiently approximate an optimal solution of the multitask learning objective by minimising a classification loss. Empirical results on two real playlist datasets show the proposed approach has good performance for cold-start playlist recommendation.
The aim of this study was to develop a digital histopathology system for identifying odontogenic keratocysts in hematoxylin- and eosin-stained tissue specimens of jaw cysts. Approximately 5000 microscopy images with 400$\times$ magnification were obtained from 199 odontogenic keratocysts, 208 dentigerous cysts, and 55 radicular cysts. A proportion of these images were used to make training patches, which were annotated as belonging to one of the following three classes: keratocysts, non-keratocysts, and stroma. The patches for the cysts contained the complete lining epithelium, with the cyst cavity being present on the upper side. The convolutional neural network (CNN) VGG16 was finetuned to this dataset. The trained CNN could recognize the basal cell palisading pattern, which is the definitive criterion for diagnosing keratocysts. Some of the remaining images were scanned and analyzed by the trained CNN, whose output was then used to train another CNN for binary classification (keratocyst or not). The area under the receiver operating characteristics curve for the entire algorithm was 0.997 for the test dataset. Thus, the proposed patch classification strategy is usable for automated keratocyst diagnosis. However, further optimization must be performed to make it suitable for practical use.
In binary classification framework, we are interested in making cost sensitive label predictions in the presence of uniform/symmetric label noise. We first observe that $0$-$1$ Bayes classifiers are not (uniform) noise robust in cost sensitive setting. To circumvent this impossibility result, we present two schemes; unlike the existing methods, our schemes do not require noise rate. The first one uses $\alpha$-weighted $\gamma$-uneven margin squared loss function, $l_{\alpha, usq}$, which can handle cost sensitivity arising due to domain requirement (using user given $\alpha$) or class imbalance (by tuning $\gamma$) or both. However, we observe that $l_{\alpha, usq}$ Bayes classifiers are also not cost sensitive and noise robust. We show that regularized ERM of this loss function over the class of linear classifiers yields a cost sensitive uniform noise robust classifier as a solution of a system of linear equations. We also provide a performance bound for this classifier. The second scheme that we propose is a re-sampling based scheme that exploits the special structure of the uniform noise models and uses in-class probability $\eta$ estimates. Our computational experiments on some UCI datasets with class imbalance show that classifiers of our two schemes are on par with the existing methods and in fact better in some cases w.r.t. Accuracy and Arithmetic Mean, without using/tuning noise rate. We also consider other cost sensitive performance measures viz., F measure and Weighted Cost for evaluation. As our re-sampling scheme requires estimates of $\eta$, we provide a detailed comparative study of various $\eta$ estimation methods on synthetic datasets, w.r.t. half a dozen evaluation criterion. Also, we provide understanding on the interpretation of cost parameters $\alpha$ and $\gamma$ using different synthetic data experiments.
Though large-scale datasets are essential for training deep learning systems, it is expensive to scale up the collection of medical imaging datasets. Synthesizing the objects of interests, such as lung nodules, in medical images based on the distribution of annotated datasets can be helpful for improving the supervised learning tasks, especially when the datasets are limited by size and class balance. In this paper, we propose the class-aware adversarial synthesis framework to synthesize lung nodules in CT images. The framework is built with a coarse-to-fine patch in-painter (generator) and two class-aware discriminators. By conditioning on the random latent variables and the target nodule labels, the trained networks are able to generate diverse nodules given the same context. By evaluating on the public LIDC-IDRI dataset, we demonstrate an example application of the proposed framework for improving the accuracy of the lung nodule malignancy estimation as a binary classification problem, which is important in the lung screening scenario. We show that combining the real image patches and the synthetic lung nodules in the training set can improve the mean AUC classification score across different network architectures by 2%.
Retrieving indexed documents, not by their topical content but their writing style opens the door for a number of applications in information retrieval (IR). One application is to retrieve textual content of a certain author X, where the queried IR system is provided beforehand with a set of reference texts of X. Authorship verification (AV), which is a research subject in the field of digital text forensics, is suitable for this purpose. The task of AV is to determine if two documents (i.e. an indexed and a reference document) have been written by the same author X. Even though AV represents a unary classification problem, a number of existing approaches consider it as a binary classification task. However, the underlying classification model of an AV method has a number of serious implications regarding its prerequisites, evaluability, and applicability. In our comprehensive literature review, we observed several misunderstandings regarding the differentiation of unary and binary AV approaches that require consideration. The objective of this paper is, therefore, to clarify these by proposing clear criteria and new properties that aim to improve the characterization of existing and future AV approaches. Given both, we investigate the applicability of eleven existing unary and binary AV methods as well as four generic unary classification algorithms on two self-compiled corpora. Furthermore, we highlight an important issue concerning the evaluation of AV methods based on fixed decision criterions, which has not been paid attention in previous AV studies.
Cosmologists are facing the problem of the analysis of a huge quantity of data when observing the sky. The methods used in cosmology are, for the most of them, relying on astrophysical models, and thus, for the classification, they usually use a machine learning approach in two-steps, which consists in, first, extracting features, and second, using a classifier. In this paper, we are specifically studying the supernovae phenomenon and especially the binary classification "I.a supernovae versus not-I.a supernovae". We present two Convolutional Neural Networks (CNNs) defeating the current state-of-the-art. The first one is adapted to time series and thus to the treatment of supernovae light-curves. The second one is based on a Siamese CNN and is suited to the nature of data, i.e. their sparsity and their weak quantity (small learning database).
In this paper, we theoretically prove that adding one special neuron per output unit eliminates all suboptimal local minima of any deep neural network, for multi-class classification, binary classification, and regression with an arbitrary loss function, under practical assumptions. At every local minimum of any deep neural network with these added neurons, the set of parameters of the original neural network (without added neurons) is guaranteed to be a global minimum of the original neural network. The effects of the added neurons are proven to automatically vanish at every local minimum. Moreover, we provide a novel theoretical characterization of a failure mode of eliminating suboptimal local minima via an additional theorem and several examples. This paper also introduces a novel proof technique based on the perturbable gradient basis (PGB) necessary condition of local minima, which provides new insight into the elimination of local minima and is applicable to analyze various models and transformations of objective functions beyond the elimination of local minima.
Event detection using social media streams needs a set of informative features with strong signals that need minimal preprocessing and are highly associated with events of interest. Identifying these informative features as keywords from Twitter is challenging, as people use informal language to express their thoughts and feelings. This informality includes acronyms, misspelled words, synonyms, transliteration and ambiguous terms. In this paper, we propose an efficient method to select the keywords frequently used in Twitter that are mostly associated with events of interest such as protests. The volume of these keywords is tracked in real time to identify the events of interest in a binary classification scheme. We use keywords within word-pairs to capture the context. The proposed method is to binarize vectors of daily counts for each word-pair by applying a spike detection temporal filter, then use the Jaccard metric to measure the similarity of the binary vector for each word-pair with the binary vector describing event occurrence. The top n word-pairs are used as features to classify any day to be an event or non-event day. The selected features are tested using multiple classifiers such as Naive Bayes, SVM, Logistic Regression, KNN and decision trees. They all produced AUC ROC scores up to 0.91 and F1 scores up to 0.79. The experiment is performed using the English language in multiple cities such as Melbourne, Sydney and Brisbane as well as the Indonesian language in Jakarta. The two experiments, comprising different languages and locations, yielded similar results.