It is well known that a systematic analysis of the pupil size variations, recorded by means of an eye-tracker, is a rich source of information about a subject's arousal and cognitive state. Current methods for pupil analysis are limited to descriptive statistics, struggle in handling the wide inter-subjects variability and must be coupled with a long series of pre-processing signal operations. In this we present a data-driven approach in which 1-D Convolutional Neural Networks are applied directly to the raw pupil size data. To test its effectiveness, we apply our method in a binary classification task with two different groups of subjects: a group of elderly patients with Parkinson disease (PDs), a condition in which pupil abnormalities have been extensively reported, and a group of healthy adults subjects (HCs). Long-range registration (10 minutes) of the pupil size were collected in scotopic conditions (complete darkness, 0 lux). 1-D convolutional neural network models are trained for classification of short-range sequences (10 to 60 seconds of registration). The model provides prediction with high average accuracy on a hold out test set. Dataset and codes are released for reproducibility and benchmarking purposes.
We provide a new online learning algorithm which utilizes online passive-aggressive learning (PA) and total-error-rate minimization (TER) for binary classification. The PA learning establishes not only large margin training but also the capacity to handle non-separable data. The TER learning on the other hand minimizes an approximated classification error based objective function. We propose an online PATER algorithm which combines those useful properties. In addition, we also present a weighted PATER algorithm to improve the ability to cope with data imbalance problems. Experimental results demonstrate that the proposed PATER algorithms achieves better performances in terms of efficiency and effectiveness than the existing state-of-the-art online learning algorithms in real-world data sets.
We consider a query-based data acquisition problem for binary classification of unknown labels, which has diverse applications in communications, crowdsourcing, recommender systems and active learning. To ensure reliable recovery of unknown labels with as few number of queries as possible, we consider an effective query type that asks "group attribute" of a chosen subset of objects. In particular, we consider the problem of classifying $m$ binary labels with XOR queries that ask whether the number of objects having a given attribute in the chosen subset of size $d$ is even or odd. The subset size $d$, which we call query degree, can be varying over queries. We consider a general noise model where the accuracy of answers on queries changes depending both on the worker (the data provider) and query degree $d$. For this general model, we characterize the information-theoretic limit on the optimal number of queries to reliably recover $m$ labels in terms of a given combination of degree-$d$ queries and noise parameters. Further, we propose an efficient inference algorithm that achieves this limit even when the noise parameters are unknown.
In this study, we analyze the role of various categories of subsidiary information in conducting replay attack spoofing detection: `Room Size', `Reverberation', `Speaker-to-ASV distance, `Attacker-to-Speaker distance', and `Replay Device Quality'. As a means of analyzing subsidiary information, we use two frameworks to either subtract or include a category of subsidiary information to the code extracted from a deep neural network. For subtraction, we utilize an adversarial process framework which makes the code orthogonal to the basis vectors of the subsidiary information. For addition, we utilize the multi-task learning framework to include subsidiary information to the code. All experiments are conducted using the ASVspoof 2019 physical access scenario with the provided meta data. Through the analysis of the result of the two approaches, we conclude that various categories of subsidiary information does not reside enough in the code when the deep neural network is trained for binary classification. Explicitly including various categories of subsidiary information through the multi-task learning framework can help improve performance in closed set condition.
Extensive empirical evidence reveals that, for a wide range of different learning methods and datasets, the risk curve exhibits a double-descent (DD) trend as a function of the model size. In a recent paper [Zeyu,Kammoun,Thrampoulidis,2019] the authors studied binary linear classification models and showed that the test error of gradient descent (GD) with logistic loss undergoes a DD. In this paper, we complement these results by extending them to GD with square loss. We show that the DD phenomenon persists, but we also identify several differences compared to logistic loss. This emphasizes that crucial features of DD curves (such as their transition threshold and global minima) depend both on the training data and on the learning algorithm. We further study the dependence of DD curves on the size of the training set. Similar to our earlier work, our results are analytic: we plot the DD curves by first deriving sharp asymptotics for the test error under Gaussian features. Albeit simple, the models permit a principled study of DD features, the outcomes of which theoretically corroborate related empirical findings occurring in more complex learning tasks.
stream-learn is a Python package compatible with scikit-learn and developed for the drifting and imbalanced data stream analysis. Its main component is a stream generator, which allows to produce a synthetic data stream that may incorporate each of the three main concept drift types (i.e. sudden, gradual and incremental drift) in their recurring or non-recurring versions. The package allows conducting experiments following established evaluation methodologies (i.e. Test-Then-Train and Prequential). In addition, estimators adapted for data stream classification have been implemented, including both simple classifiers and state-of-art chunk-based and online classifier ensembles. To improve computational efficiency, package utilises its own implementations of prediction metrics for imbalanced binary classification tasks.
Positive-confidence (Pconf) classification [Ishida et al., 2018] is a promising weakly-supervised learning method which trains a binary classifier only from positive data equipped with confidence. However, in practice, the confidence may be skewed by bias arising in an annotation process. The Pconf classifier cannot be properly learned with skewed confidence, and consequently, the classification performance might be deteriorated. In this paper, we introduce the parameterized model of the skewed confidence, and propose the method for selecting the hyperparameter which cancels out the negative impact of skewed confidence under the assumption that we have the misclassification rate of positive samples as a prior knowledge. We demonstrate the effectiveness of the proposed method through a synthetic experiment with simple linear models and benchmark problems with neural network models. We also apply our method to drivers' drowsiness prediction to show that it works well with a real-world problem where confidence is obtained based on manual annotation.
In this paper, we introduce a neural network framework for semi-supervised clustering (SSC) with pairwise (must-link or cannot-link) constraints. In contrast to existing approaches, we decompose SSC into two simpler classification tasks/stages: the first stage uses a pair of Siamese neural networks to label the unlabeled pairs of points as must-link or cannot-link; the second stage uses the fully pairwise-labeled dataset produced by the first stage in a supervised neural-network-based clustering method. The proposed approach, S3C2 (Semi-Supervised Siamese Classifiers for Clustering), is motivated by the observation that binary classification (such as assigning pairwise relations) is usually easier than multi-class clustering with partial supervision. On the other hand, being classification-based, our method solves only well-defined classification problems, rather than less well specified clustering tasks. Extensive experiments on various datasets demonstrate the high performance of the proposed method.
Aggregated hold-out (Agghoo) is a method which averages learning rules selected by hold-out (that is, cross-validation with a single split). We provide the first theoretical guarantees on Agghoo, ensuring that it can be used safely: Agghoo performs at worst like the hold-out when the risk is convex. The same holds true in classification with the 0-1 risk, with an additional constant factor. For the hold-out, oracle inequalities are known for bounded losses, as in binary classification. We show that similar results can be proved, under appropriate assumptions, for other risk-minimization problems. In particular, we obtain an oracle inequality for regularized kernel regression with a Lip-schitz loss, without requiring that the Y variable or the regressors be bounded. Numerical experiments show that aggregation brings a significant improvement over the hold-out and that Agghoo is competitive with cross-validation.
Ensemble learning is a powerful paradigm that has been usedin the top state-of-the-art machine learning methods like Random Forestsand XGBoost. Inspired by the success of such methods, we have devel-oped a new Genetic Programming method called Ensemble GP. The evo-lutionary cycle of Ensemble GP follows the same steps as other GeneticProgramming systems, but with differences in the population structure,fitness evaluation and genetic operators. We have tested this method oneight binary classification problems, achieving results significantly betterthan standard GP, with much smaller models. Although other methodslike M3GP and XGBoost were the best overall, Ensemble GP was able toachieve exceptionally good generalization results on a particularly hardproblem where none of the other methods was able to succeed.
The goal of the change-point detection is to discover changes of time series distribution. One of the state of the art approaches of the change-point detection are based on direct density ratio estimation. In this work we show how existing algorithms can be generalized using various binary classification and regression models. In particular, we show that the Gradient Boosting over Decision Trees and Neural Networks can be used for this purpose. The algorithms are tested on several synthetic and real-world datasets. The results show that the proposed methods outperform classical RuLSIF algorithm. Discussion of cases where the proposed algorithms have advantages over existing methods are also provided.
Words and phrases acquire meaning from the way they are used in society, from their relative semantics to other words and phrases. For computers the equivalent of `society' is `database,' and the equivalent of `use' is `way to search the database.' We present a new theory of similarity between words and phrases based on information distance and Kolmogorov complexity. To fix thoughts we use the world-wide-web as database, and Google as search engine. The method is also applicable to other search engines and databases. This theory is then applied to construct a method to automatically extract similarity, the Google similarity distance, of words and phrases from the world-wide-web using Google page counts. The world-wide-web is the largest database on earth, and the context information entered by millions of independent users averages out to provide automatic semantics of useful quality. We give applications in hierarchical clustering, classification, and language translation. We give examples to distinguish between colors and numbers, cluster names of paintings by 17th century Dutch masters and names of books by English novelists, the ability to understand emergencies, and primes, and we demonstrate the ability to do a simple automatic English-Spanish translation. Finally, we use the WordNet database as an objective baseline against which to judge the performance of our method. We conduct a massive randomized trial in binary classification using support vector machines to learn categories based on our Google distance, resulting in an a mean agreement of 87% with the expert crafted WordNet categories.
Supervised learning requires a sufficient training dataset which includes all label. However, there are cases that some class is not in the training data. Zero-Shot Learning (ZSL) is the task of predicting class that is not in the training data(target class). The existing ZSL method is done for image data. However, the zero-shot problem should happen to every data type. Hence, considering ZSL for other data types is required. In this paper, we propose the cluster-based ZSL method, which is a baseline method for multivariate binary classification problems. The proposed method is based on the assumption that if data is far from training data, the data is considered as target class. In training, clustering is done for training data. In prediction, the data is determined belonging to a cluster or not. If data does not belong to a cluster, the data is predicted as target class. The proposed method is evaluated and demonstrated using the KEEL dataset. This paper has been published in the Journal of Ambient Intelligence and Humanized Computing. The final version is available at the following URL: https://link.springer.com/article/10.1007/s12652-020-02268-5
Deep learning is also known as hierarchical learning, where the learner _learns_ to represent a complicated target function by decomposing it into a sequence of simpler functions to reduce sample and time complexity. This paper formally analyzes how multi-layer neural networks can perform such hierarchical learning _efficiently_ and _automatically_ by SGD on the training objective. On the conceptual side, we present a theoretical characterizations of how certain types of deep (i.e. super-constant layer) neural networks can still be sample and time efficiently trained on some hierarchical tasks, when no existing algorithm (including layerwise training, kernel method, etc) is known to be efficient. We establish a new principle called "backward feature correction", where the errors in the lower-level features can be automatically corrected when training together with the higher-level layers. We believe this is a key behind how deep learning is performing deep (hierarchical) learning, as opposed to layerwise learning or simulating some non-hierarchical method. On the technical side, we show for every input dimension $d > 0$, there is a concept class of degree $\omega(1)$ multi-variate polynomials so that, using $\omega(1)$-layer neural networks as learners, SGD can learn any function from this class in $\mathsf{poly}(d)$ time to any $\frac{1}{\mathsf{poly}(d)}$ error, through learning to represent it as a composition of $\omega(1)$ layers of quadratic functions using "backward feature correction." In contrast, we do not know any other simpler algorithm (including layerwise training, applying kernel method sequentially, training a two-layer network, etc) that can learn this concept class in $\mathsf{poly}(d)$ time even to any $d^{-0.01}$ error. As a side result, we prove $d^{\omega(1)}$ lower bounds for several non-hierarchical learners, including any kernel methods.
Semi-supervised learning is a model training method that uses both labeled and unlabeled data. This paper proposes a fully Bayes semi-supervised learning algorithm that can be applied to any multi-category classification problem. We assume the labels are missing at random when using unlabeled data in a semi-supervised setting. Suppose we have $K$ classes in the data. We assume that the observations follow $K$ multivariate normal distributions depending on their true class labels after some common unknown transformation is applied to each component of the observation vector. The function is expanded in a B-splines series, and a prior is added to the coefficients. We consider a normal prior on the coefficients and constrain the values to meet the normality and identifiability constraints requirement. The precision matrices of the Gaussian distributions are given a conjugate Wishart prior, while the means are given the improper uniform prior. The resulting posterior is still conditionally conjugate, and the Gibbs sampler aided by a data-augmentation technique can thus be adopted. An extensive simulation study compares the proposed method with several other available methods. The proposed method is also applied to real datasets on diagnosing breast cancer and classification of signals. We conclude that the proposed method has a better prediction accuracy in various cases.
Fungal Biosynthetic Gene Clusters (BGCs) of secondary metabolites are clusters of genes capable of producing natural products, compounds that play an important role in the production of a wide variety of bioactive compounds, including antibiotics and pharmaceuticals. Identifying BGCs can lead to the discovery of novel natural products to benefit human health. Previous work has been focused on developing automatic tools to support BGC discovery in plants, fungi, and bacteria. Data-driven methods, as well as probabilistic and supervised learning methods have been explored in identifying BGCs. Most methods applied to identify fungal BGCs were data-driven and presented limited scope. Supervised learning methods have been shown to perform well at identifying BGCs in bacteria, and could be well suited to perform the same task in fungi. But labeled data instances are needed to perform supervised learning. Openly accessible BGC databases contain only a very small portion of previously curated fungal BGCs. Making new fungal BGC datasets available could motivate the development of supervised learning methods for fungal BGCs and potentially improve prediction performance compared to data-driven methods. In this work we propose new publicly available fungal BGC datasets to support the BGC discovery task using supervised learning. These datasets are prepared to perform binary classification and predict candidate BGC regions in fungal genomes. In addition we analyse the performance of a well supported supervised learning tool developed to predict BGCs.
The correct use of Dutch pronouns 'die' and 'dat' is a stumbling block for both native and non-native speakers of Dutch due to the multiplicity of syntactic functions and the dependency on the antecedent's gender and number. Drawing on previous research conducted on neural context-dependent dt-mistake correction models (Heyman et al. 2018), this study constructs the first neural network model for Dutch demonstrative and relative pronoun resolution that specifically focuses on the correction and part-of-speech prediction of these two pronouns. Two separate datasets are built with sentences obtained from, respectively, the Dutch Europarl corpus (Koehn 2015) - which contains the proceedings of the European Parliament from 1996 to the present - and the SoNaR corpus (Oostdijk et al. 2013) - which contains Dutch texts from a variety of domains such as newspapers, blogs and legal texts. Firstly, a binary classification model solely predicts the correct 'die' or 'dat'. The classifier with a bidirectional long short-term memory architecture achieves 84.56% accuracy. Secondly, a multitask classification model simultaneously predicts the correct 'die' or 'dat' and its part-of-speech tag. The model containing a combination of a sentence and context encoder with both a bidirectional long short-term memory architecture results in 88.63% accuracy for die/dat prediction and 87.73% accuracy for part-of-speech prediction. More evenly-balanced data, larger word embeddings, an extra bidirectional long short-term memory layer and integrated part-of-speech knowledge positively affects die/dat prediction performance, while a context encoder architecture raises part-of-speech prediction performance. This study shows promising results and can serve as a starting point for future research on machine learning models for Dutch anaphora resolution.
Transfer learning --- transferring learned knowledge --- has brought a paradigm shift in the way models are trained. The lucrative benefits of improved accuracy and reduced training time have shown promise in training models with constrained computational resources and fewer training samples. Specifically, publicly available text-based models such as GloVe and BERT that are trained on large corpus of datasets have seen ubiquitous adoption in practice. In this paper, we ask, "can transfer learning in text prediction models be exploited to perform misclassification attacks?" As our main contribution, we present novel attack techniques that utilize unintended features learnt in the teacher (public) model to generate adversarial examples for student (downstream) models. To the best of our knowledge, ours is the first work to show that transfer learning from state-of-the-art word-based and sentence-based teacher models increase the susceptibility of student models to misclassification attacks. First, we propose a novel word-score based attack algorithm for generating adversarial examples against student models trained using context-free word-level embedding model. On binary classification tasks trained using the GloVe teacher model, we achieve an average attack accuracy of 97% for the IMDB Movie Reviews and 80% for the Fake News Detection. For multi-class tasks, we divide the Newsgroup dataset into 6 and 20 classes and achieve an average attack accuracy of 75% and 41% respectively. Next, we present length-based and sentence-based misclassification attacks for the Fake News Detection task trained using a context-aware BERT model and achieve 78% and 39% attack accuracy respectively. Thus, our results motivate the need for designing training techniques that are robust to unintended feature learning, specifically for transfer learned models.
Fine-grained software vulnerability detection is an important and challenging problem. Ideally, a detection system (or detector) not only should be able to detect whether or not a program contains vulnerabilities, but also should be able to pinpoint the type of a vulnerability in question. Existing vulnerability detection methods based on deep learning can detect the presence of vulnerabilities (i.e., addressing the binary classification or detection problem), but cannot pinpoint types of vulnerabilities (i.e., incapable of addressing multiclass classification). In this paper, we propose the first deep learning-based system for multiclass vulnerability detection, dubbed $\mu$VulDeePecker. The key insight underlying $\mu$VulDeePecker is the concept of code attention, which can capture information that can help pinpoint types of vulnerabilities, even when the samples are small. For this purpose, we create a dataset from scratch and use it to evaluate the effectiveness of $\mu$VulDeePecker. Experimental results show that $\mu$VulDeePecker is effective for multiclass vulnerability detection and that accommodating control-dependence (other than data-dependence) can lead to higher detection capabilities.
We conduct a post hoc analysis of solar flare predictions made by a Long Short Term Memory (LSTM) model employing data in the form of Space-weather HMI Active Region Patches (SHARP) parameters calculated from data in proximity to the magnetic polarity inversion line where the flares originate. We train the the LSTM model for binary classification to provide a prediction score for the probability of M/X class flares to occur in next hour. We then develop a dimension-reduction technique to reduce the dimensions of SHARP parameter (LSTM inputs) and demonstrate the different patterns of SHARP parameters corresponding to the transition from low to high prediction score. Our work shows that a subset of SHARP parameters contain the key signals that strong solar flare eruptions are imminent. The dynamics of these parameters have a highly uniform trajectory for many events whose LSTM prediction scores for M/X class flares transition from very low to very high. The results demonstrate the existence of a few threshold values of SHARP parameters that when surpassed indicate a high probability of the eruption of a strong flare. Our method has distilled the knowledge of solar flare eruption learnt by deep learning model and provides a more interpretable approximation, which provides physical insight to processes driving solar flares.