Hierarchical Classification (HC) is a supervised learning problem where unlabeled instances are classified into a taxonomy of classes. Several methods that utilize the hierarchical structure have been developed to improve the HC performance. However, in most cases apriori defined hierarchical structure by domain experts is inconsistent; as a consequence performance improvement is not noticeable in comparison to flat classification methods. We propose a scalable data-driven filter based rewiring approach to modify an expert-defined hierarchy. Experimental comparisons of top-down HC with our modified hierarchy, on a wide range of datasets shows classification performance improvement over the baseline hierarchy (i:e:, defined by expert), clustered hierarchy and flattening based hierarchy modification approaches. In comparison to existing rewiring approaches, our developed method (rewHier) is computationally efficient, enabling it to scale to datasets with large numbers of classes, instances and features. We also show that our modified hierarchy leads to improved classification performance for classes with few training samples in comparison to flat and state-of-the-art HC approaches.
We investigate metric learning in the context of dynamic time warping (DTW), the by far most popular dissimilarity measure used for the comparison and analysis of motion capture data. While metric learning enables a problem-adapted representation of data, the majority of methods has been proposed for vectorial data only. In this contribution, we extend the popular principle offered by the large margin nearest neighbors learner (LMNN) to DTW by treating the resulting component-wise dissimilarity values as features. We demonstrate that this principle greatly enhances the classification accuracy in several benchmarks. Further, we show that recent auxiliary concepts such as metric regularization can be transferred from the vectorial case to component-wise DTW in a similar way. We illustrate that metric regularization constitutes a crucial prerequisite for the interpretation of the resulting relevance profiles.
Crowdsourcing has been successfully employed in the past as an effective and cheap way to execute classification tasks and has therefore attracted the attention of the research community. However, we still lack a theoretical understanding of how to collect the labels from the crowd in an optimal way. In this paper we focus on the problem of worker allocation and compare two active learning policies proposed in the empirical literature with a uniform allocation of the available budget. To this end we make a thorough mathematical analysis of the problem and derive a new bound on the performance of the system. Furthermore we run extensive simulations in a more realistic scenario and show that our theoretical results hold in practice.
This paper analyzes k nearest neighbor classification with training data anonymized using anatomy. Anatomy preserves all data values, but introduces uncertainty in the mapping between identifying and sensitive values. We first study the theoretical effect of the anatomized training data on the k nearest neighbor error rate bounds, nearest neighbor convergence rate, and Bayesian error. We then validate the derived bounds empirically. We show that 1) Learning from anatomized data approaches the limits of learning through the unprotected data (although requiring larger training data), and 2) nearest neighbor using anatomized data outperforms nearest neighbor on generalization-based anonymization.
Based on API call sequences, semantic-aware and machine learning (ML) based malware classifiers can be built for malware detection or classification. Previous works concentrate on crafting and extracting various features from malware binaries, disassembled binaries or API calls via static or dynamic analysis and resorting to ML to build classifiers. However, they tend to involve too much feature engineering and fail to provide interpretability. We solve these two problems with the recent advances in deep learning: 1) RNN-based autoencoders (RNN-AEs) can automatically learn low-dimensional representation of a malware from its raw API call sequence. 2) Multiple decoders can be trained under different supervisions to give more information, other than the class or family label of a malware. Inspired by the works of document classification and automatic sentence summarization, each API call sequence can be regarded as a sentence. In this paper, we make the first attempt to build a multi-task malware learning model based on API call sequences. The model consists of two decoders, one for malware classification and one for $\emph{file access pattern}$ (FAP) generation given the API call sequence of a malware. We base our model on the general seq2seq framework. Experiments show that our model can give competitive classification results as well as insightful FAP information.
Semantic segmentation is challenging as it requires both object-level information and pixel-level accuracy. Recently, FCN-based systems gained great improvement in this area. Unlike classification networks, combining features of different layers plays an important role in these dense prediction models, as these features contains information of different levels. A number of models have been proposed to show how to use these features. However, what is the best architecture to make use of features of different layers is still a question. In this paper, we propose a module, called mixed context network, and show that our presented system outperforms most existing semantic segmentation systems by making use of this module.
Corporations are retaining ever-larger corpuses of personal data; the frequency or breaches and corresponding privacy impact have been rising accordingly. One way to mitigate this risk is through use of anonymized data, limiting the exposure of individual data to only where it is absolutely needed. This would seem particularly appropriate for data mining, where the goal is generalizable knowledge rather than data on specific individuals. In practice, corporate data miners often insist on original data, for fear that they might "miss something" with anonymized or differentially private approaches. This paper provides a theoretical justification for the use of anonymized data. Specifically, we show that a support vector classifier trained on anatomized data satisfying l-diversity should be expected to do as well as on the original data. Anatomy preserves all data values, but introduces uncertainty in the mapping between identifying and sensitive values, thus satisfying l-diversity. The theoretical effectiveness of the proposed approach is validated using several publicly available datasets, showing that we outperform the state of the art for support vector classification using training data protected by k-anonymity, and are comparable to learning on the original data.
This paper proposes a client-server decision tree learning method for outsourced private data. The privacy model is anatomization/fragmentation: the server sees data values, but the link between sensitive and identifying information is encrypted with a key known only to clients. Clients have limited processing and storage capability. Both sensitive and identifying information thus are stored on the server. The approach presented also retains most processing at the server, and client-side processing is amortized over predictions made by the clients. Experiments on various datasets show that the method produces decision trees approaching the accuracy of a non-private decision tree, while substantially reducing the client's computing resource requirements.
Short internet video clips like vines present a significantly wild distribution compared to traditional video datasets. In this paper, we focus on the problem of unsupervised action classification in wild vines using traditional labeled datasets. To this end, we use a data augmentation based simple domain adaptation strategy. We utilise semantic word2vec space as a common subspace to embed video features from both, labeled source domain and unlablled target domain. Our method incrementally augments the labeled source with target samples and iteratively modifies the embedding function to bring the source and target distributions together. Additionally, we utilise a multi-modal representation that incorporates noisy semantic information available in form of hash-tags. We show the effectiveness of this simple adaptation technique on a test set of vines and achieve notable improvements in performance.
The aim of this work is to classify the aerospace structure defects detected by eddy current non-destructive testing. The proposed method is based on the assumption that the defect is bound to the reaction of the probe coil impedance during the test. Impedance plane analysis is used to extract a feature vector from the shape of the coil impedance in the complex plane, through the use of some geometric parameters. Shape recognition is tested with three different machine-learning based classifiers: decision trees, neural networks and Naive Bayes. The performance of the proposed detection system are measured in terms of accuracy, sensitivity, specificity, precision and Matthews correlation coefficient. Several experiments are performed on dataset of eddy current signal samples for aircraft structures. The obtained results demonstrate the usefulness of our approach and the competiveness against existing descriptors.
Linear Discriminant Analysis (LDA) is a well-known technique for feature extraction and dimension reduction. The performance of classical LDA, however, significantly degrades on the High Dimension Low Sample Size (HDLSS) data for the ill-posed inverse problem. Existing approaches for HDLSS data classification typically assume the data in question are with Gaussian distribution and deal the HDLSS classification problem with regularization. However, these assumptions are too strict to hold in many emerging real-life applications, such as enabling personalized predictive analysis using Electronic Health Records (EHRs) data collected from an extremely limited number of patients who have been diagnosed with or without the target disease for prediction. In this paper, we revised the problem of predictive analysis of disease using personal EHR data and LDA classifier. To fill the gap, in this paper, we first studied an analytical model that understands the accuracy of LDA for classifying data with arbitrary distribution. The model gives a theoretical upper bound of LDA error rate that is controlled by two factors: (1) the statistical convergence rate of (inverse) covariance matrix estimators and (2) the divergence of the training/testing datasets to fitted distributions. To this end, we could lower the error rate by balancing the two factors for better classification performance. Hereby, we further proposed a novel LDA classifier De-Sparse that leverages De-sparsified Graphical Lasso to improve the estimation of LDA, which outperforms state-of-the-art LDA approaches developed for HDLSS data. Such advances and effectiveness are further demonstrated by both theoretical analysis and extensive experiments on EHR datasets.
Deep neural network (DNN) models have recently obtained state-of-the-art prediction accuracy for the transcription factor binding (TFBS) site classification task. However, it remains unclear how these approaches identify meaningful DNA sequence signals and give insights as to why TFs bind to certain locations. In this paper, we propose a toolkit called the Deep Motif Dashboard (DeMo Dashboard) which provides a suite of visualization strategies to extract motifs, or sequence patterns from deep neural network models for TFBS classification. We demonstrate how to visualize and understand three important DNN models: convolutional, recurrent, and convolutional-recurrent networks. Our first visualization method is finding a test sequence's saliency map which uses first-order derivatives to describe the importance of each nucleotide in making the final prediction. Second, considering recurrent models make predictions in a temporal manner (from one end of a TFBS sequence to the other), we introduce temporal output scores, indicating the prediction score of a model over time for a sequential input. Lastly, a class-specific visualization strategy finds the optimal input sequence for a given TFBS positive class via stochastic gradient optimization. Our experimental results indicate that a convolutional-recurrent architecture performs the best among the three architectures. The visualization techniques indicate that CNN-RNN makes predictions by modeling both motifs as well as dependencies among them.
Generating user interpretable multi-class predictions in data rich environments with many classes and explanatory covariates is a daunting task. We introduce Diagonal Orthant Latent Dirichlet Allocation (DOLDA), a supervised topic model for multi-class classification that can handle both many classes as well as many covariates. To handle many classes we use the recently proposed Diagonal Orthant (DO) probit model (Johndrow et al., 2013) together with an efficient Horseshoe prior for variable selection/shrinkage (Carvalho et al., 2010). We propose a computationally efficient parallel Gibbs sampler for the new model. An important advantage of DOLDA is that learned topics are directly connected to individual classes without the need for a reference class. We evaluate the model's predictive accuracy on two datasets and demonstrate DOLDA's advantage in interpreting the generated predictions.
We develop a class of rules spanning the range between quadratic discriminant analysis and naive Bayes, through a path of sparse graphical models. A group lasso penalty is used to introduce shrinkage and encourage a similar pattern of sparsity across precision matrices. It gives sparse estimates of interactions and produces interpretable models. Inspired by the connected-components structure of the estimated precision matrices, we propose the community Bayes model, which partitions features into several conditional independent communities and splits the classification problem into separate smaller ones. The community Bayes idea is quite general and can be applied to non-Gaussian data and likelihood-based classifiers.
The hybrid clustering-classification neural network is proposed. This network allows increasing a quality of information processing under the condition of overlapping classes due to the rational choice of a learning rate parameter and introducing a special procedure of fuzzy reasoning in the clustering process, which occurs both with an external learning signal (supervised) and without the one (unsupervised). As similarity measure neighborhood function or membership one, cosine structures are used, which allow to provide a high flexibility due to self-learning-learning process and to provide some new useful properties. Many realized experiments have confirmed the efficiency of proposed hybrid clustering-classification neural network; also, this network was used for solving diagnostics task of reactive arthritis.
Kernel normalization methods have been employed to improve robustness of optimization methods to reparametrization of convolution kernels, covariate shift, and to accelerate training of Convolutional Neural Networks (CNNs). However, our understanding of theoretical properties of these methods has lagged behind their success in applications. We develop a geometric framework to elucidate underlying mechanisms of a diverse range of kernel normalization methods. Our framework enables us to expound and identify geometry of space of normalized kernels. We analyze and delineate how state-of-the-art kernel normalization methods affect the geometry of search spaces of the stochastic gradient descent (SGD) algorithms in CNNs. Following our theoretical results, we propose a SGD algorithm with assurance of almost sure convergence of the methods to a solution at single minimum of classification loss of CNNs. Experimental results show that the proposed method achieves state-of-the-art performance for major image classification benchmarks with CNNs.
In classification problems when multiples algorithms are applied to different benchmarks a difficult issue arises, i.e., how can we rank the algorithms? In machine learning it is common run the algorithms several times and then a statistic is calculated in terms of means and standard deviations. In order to compare the performance of the algorithms, it is very common to employ statistical tests. However, these tests may also present limitations, since they consider only the means and not the standard deviations of the obtained results. In this paper, we present the so called A-TOPSIS, based on TOPSIS (Technique for Order Preference by Similarity to Ideal Solution), to solve the problem of ranking and comparing classification algorithms in terms of means and standard deviations. We use two case studies to illustrate the A-TOPSIS for ranking classification algorithms and the results show the suitability of A-TOPSIS to rank the algorithms. The presented approach is general and can be applied to compare the performance of stochastic algorithms in machine learning. Finally, to encourage researchers to use the A-TOPSIS for ranking algorithms we also presented in this work an easy-to-use A-TOPSIS web framework.
In this paper, we propose and compare two spectral angle based approaches for spatial-spectral classification. Our methods use the spectral angle to generate unary energies in a grid-structured Markov random field defined over the pixel labels of a hyperspectral image. The first approach is to use the exponential spectral angle mapper (ESAM) kernel/covariance function, a spectral angle based function, with the support vector machine and the Gaussian process classifier. The second approach is to directly use the minimum spectral angle between the test pixel and the training pixels as the unary energy. We compare the proposed methods with the state-of-the-art Markov random field methods that use support vector machines and Gaussian processes with squared exponential kernel/covariance function. In our experiments with two datasets, it is seen that using minimum spectral angle as unary energy produces better or comparable results to the existing methods at a smaller running time.
In this paper, we introduce a novel fusion method that can enhance object detection performance by fusing decisions from two different types of computer vision tasks: object detection and image classification. In the proposed work, the class label of an image obtained from image classification is viewed as prior knowledge about existence or non-existence of certain objects. The prior knowledge is then fused with the decisions of object detection to improve detection accuracy by mitigating false positives of an object detector that are strongly contradicted with the prior knowledge. A recently introduced novel fusion approach called dynamic belief fusion (DBF) is used to fuse the detector output with the classification prior. Experimental results show that the detection performance of all the detection algorithms used in the proposed work is improved on benchmark datasets via the proposed fusion framework.
In recent years, traditional cybersecurity safeguards have proven ineffective against insider threats. Famous cases of sensitive information leaks caused by insiders, including the WikiLeaks release of diplomatic cables and the Edward Snowden incident, have greatly harmed the U.S. government's relationship with other governments and with its own citizens. Data Leak Prevention (DLP) is a solution for detecting and preventing information leaks from within an organization's network. However, state-of-art DLP detection models are only able to detect very limited types of sensitive information, and research in the field has been hindered due to the lack of available sensitive texts. Many researchers have focused on document-based detection with artificially labeled "confidential documents" for which security labels are assigned to the entire document, when in reality only a portion of the document is sensitive. This type of whole-document based security labeling increases the chances of preventing authorized users from accessing non-sensitive information within sensitive documents. In this paper, we introduce Automated Classification Enabled by Security Similarity (ACESS), a new and innovative detection model that penetrates the complexity of big text security classification/detection. To analyze the ACESS system, we constructed a novel dataset, containing formerly classified paragraphs from diplomatic cables made public by the WikiLeaks organization. To our knowledge this paper is the first to analyze a dataset that contains actual formerly sensitive information annotated at paragraph granularity.