The volume of data being collected in solar physics has exponentially increased over the past decade and with the introduction of the $\textit{Daniel K. Inouye Solar Telescope}$ (DKIST) we will be entering the age of petabyte solar data. Automated feature detection will be an invaluable tool for post-processing of solar images to create catalogues of data ready for researchers to use. We propose a deep learning model to accomplish this; a deep convolutional neural network is adept at feature extraction and processing images quickly. We train our network using data from $\textit{Hinode/Solar Optical Telescope}$ (SOT) H$\alpha$ images of a small subset of solar features with different geometries: filaments, prominences, flare ribbons, sunspots and the quiet Sun ($\textit{i.e.}$ the absence of any of the other four features). We achieve near perfect performance on classifying unseen images from SOT ($\approx$99.9\%) in 4.66 seconds. We also for the first time explore transfer learning in a solar context. Transfer learning uses pre-trained deep neural networks to help train new deep learning models $\textit{i.e.}$ it teaches a new model. We show that our network is robust to changes in resolution by degrading images from SOT resolution ($\approx$0.33$^{\prime \prime}$ at $\lambda$=6563\AA{}) to $\textit{Solar Dynamics Observatory/Atmospheric Imaging Assembly}$ (SDO/AIA) resolution ($\approx$1.2$^{\prime \prime}$) without a change in performance of our network. However, we also observe where the network fails to generalise to sunspots from SDO/AIA bands 1600/1700\AA{} due to small-scale brightenings around the sunspots and prominences in SDO/AIA 304\AA{} due to coronal emission.
Recently, many works have been inspired by the success of deep learning in computer vision for plant diseases classification. Unfortunately, these end-to-end deep classifiers lack transparency which can limit their adoption in practice. In this paper, we propose a new trainable visualization method for plant diseases classification based on a Convolutional Neural Network (CNN) architecture composed of two deep classifiers. The first one is named Teacher and the second one Student. This architecture leverages the multitask learning to train the Teacher and the Student jointly. Then, the communicated representation between the Teacher and the Student is used as a proxy to visualize the most important image regions for classification. This new architecture produces sharper visualization than the existing methods in plant diseases context. All experiments are achieved on PlantVillage dataset that contains 54306 plant images.
Ensemble approaches for uncertainty estimation have recently been applied to the tasks of misclassification detection, out-of-distribution input detection and adversarial attack detection. Prior Networks have been proposed as an approach to efficiently \emph{emulate} an ensemble of models for classification by parameterising a Dirichlet prior distribution over output distributions. These models have been shown to outperform alternative ensemble approaches, such as Monte-Carlo Dropout, on the task of out-of-distribution input detection. However, scaling Prior Networks to complex datasets with many classes is difficult using the training criteria originally proposed. This paper makes two contributions. First, we show that the appropriate training criterion for Prior Networks is the \emph{reverse} KL-divergence between Dirichlet distributions. This addresses issues in the nature of the training data target distributions, enabling prior networks to be successfully trained on classification tasks with arbitrarily many classes, as well as improving out-of-distribution detection performance. Second, taking advantage of this new training criterion, this paper investigates using Prior Networks to detect adversarial attacks and proposes a generalized form of adversarial training. It is shown that the construction of successful \emph{adaptive} whitebox attacks, which affect the prediction and evade detection, against Prior Networks trained on CIFAR-10 and CIFAR-100 using the proposed approach requires a greater amount of computational effort than against networks defended using standard adversarial training or MC-dropout.
Open information extraction (IE) is the task of extracting open-domain assertions from natural language sentences. A key step in open IE is confidence modeling, ranking the extractions based on their estimated quality to adjust precision and recall of extracted assertions. We found that the extraction likelihood, a confidence measure used by current supervised open IE systems, is not well calibrated when comparing the quality of assertions extracted from different sentences. We propose an additional binary classification loss to calibrate the likelihood to make it more globally comparable, and an iterative learning process, where extractions generated by the open IE model are incrementally included as training samples to help the model learn from trial and error. Experiments on OIE2016 demonstrate the effectiveness of our method. Code and data are available at https://github.com/jzbjyb/oie_rank.
As the application of deep neural networks proliferates in numerous areas such as medical imaging, video surveillance, and self driving cars, the need for explaining the decisions of these models has become a hot research topic, both at the global and local level. Locally, most explanation methods have focused on identifying relevance of features, limiting the types of explanations possible. In this paper, we investigate a new direction by leveraging latent features to generate contrastive explanations; predictions are explained not only by highlighting aspects that are in themselves sufficient to justify the classification, but also by new aspects which if added will change the classification. The key contribution of this paper lies in how we add features to rich data in a formal yet humanly interpretable way that leads to meaningful results. Our new definition of "addition" uses latent features to move beyond the limitations of previous explanations and resolve an open question laid out in Dhurandhar, et. al. (2018), which creates local contrastive explanations but is limited to simple datasets such as grayscale images. The strength of our approach in creating intuitive explanations that are also quantitatively superior to other methods is demonstrated on three diverse image datasets (skin lesions, faces, and fashion apparel). A user study with 200 participants further exemplifies the benefits of contrastive information, which can be viewed as complementary to other state-of-the-art interpretability methods.
This work studies the denoising of piecewise smooth graph signals that exhibit inhomogeneous levels of smoothness over a graph, where the value at each node can be vector-valued. We extend the graph trend filtering framework to denoising vector-valued graph signals with a family of non-convex regularizers, which exhibit superior recovery performance over existing convex regularizers. Using an oracle inequality, we establish the statistical error rates of first-order stationary points of the proposed non-convex method for generic graphs. Furthermore, we present an ADMM-based algorithm to solve the proposed method and establish its convergence. Numerical experiments are conducted on both synthetic and real-world data for denoising, support recovery, event detection, and semi-supervised classification.
Recently, graph neural networks (GNNs) have proved to be suitable in tasks on unstructured data. Particularly in tasks as community detection, node classification, and link prediction. However, most GNN models still operate with static relationships. We propose the Graph Learning Network (GLN), a simple yet effective process to learn node embeddings and structure prediction functions. Our model uses graph convolutions to propose expected node features, and predict the best structure based on them. We repeat these steps recursively to enhance the prediction and the embeddings.
We show that the classification performance of graph convolutional networks (GCNs) is related to the alignment between features, graph, and ground truth, which we quantify using a subspace alignment measure (SAM) corresponding to the Frobenius norm of the matrix of pairwise chordal distances between three subspaces associated with features, graph, and ground truth. The proposed measure is based on the principal angles between subspaces and has both spectral and geometrical interpretations. We showcase the relationship between the SAM and the classification performance through the study of limiting cases of GCNs and systematic randomizations of both features and graph structure applied to a constructive example and several examples of citation networks of different origins. The analysis also reveals the relative importance of the graph and features for classification purposes.
Recurrent neural networks (RNNs) have been extraordinarily successful for prediction with sequential data. To tackle highly variable and noisy real-world data, we introduce Particle Filter Recurrent Neural Networks (PF-RNNs), a new RNN family that explicitly models uncertainty in its internal structure: while an RNN relies on a long, deterministic latent state vector, a PF-RNN maintains a latent state distribution, approximated as a set of particles. For effective learning, we provide a fully differentiable particle filter algorithm that updates the PF-RNN latent state distribution according to the Bayes rule. Experiments demonstrate that the proposed PF-RNNs outperform the corresponding standard gated RNNs on a synthetic robot localization dataset and 10 real-world sequence prediction datasets for text classification, stock price prediction, etc.
Estimating the predictive uncertainty of a Bayesian learning model is critical in various decision-making problems, e.g., reinforcement learning, detecting adversarial attack, self-driving car. As the model posterior is almost always intractable, most efforts were made on finding an accurate approximation the true posterior. Even though a decent estimation of the model posterior is obtained, another approximation is required to compute the predictive distribution over the desired output. A common accurate solution is to use Monte Carlo (MC) integration. However, it needs to maintain a large number of samples, evaluate the model repeatedly and average multiple model outputs. In many real-world cases, this is computationally prohibitive. In this work, assuming that the exact posterior or a decent approximation is obtained, we propose a generic framework to approximate the output probability distribution induced by model posterior with a parameterized model and in an amortized fashion. The aim is to approximate the true uncertainty of a specific Bayesian model, meanwhile alleviating the heavy workload of MC integration at testing time. The proposed method is universally applicable to Bayesian classification models that allow for posterior sampling. Theoretically, we show that the idea of amortization incurs no additional costs on approximation performance. Empirical results validate the strong practical performance of our approach.
Differential privacy (DP) is a popular mechanism for training machine learning models with bounded leakage about the presence of specific points in the training data. The cost of differential privacy is a reduction in the model's accuracy. We demonstrate that in the neural networks trained using differentially private stochastic gradient descent (DP-SGD), this cost is not borne equally: accuracy of DP models drops much more for the underrepresented classes and subgroups. For example, a gender classification model trained using DP-SGD exhibits much lower accuracy for black faces than for white faces. Critically, this gap is bigger in the DP model than in the non-DP model, i.e., if the original model is unfair, the unfairness becomes worse once DP is applied. We demonstrate this effect for a variety of tasks and models, including sentiment analysis of text and image classification. We then explain why DP training mechanisms such as gradient clipping and noise addition have disproportionate effect on the underrepresented and more complex subgroups, resulting in a disparate reduction of model accuracy.
In federated learning problems, data is scattered across different servers and exchanging or pooling it is often impractical or prohibited. We develop a Bayesian nonparametric framework for federated learning with neural networks. Each data server is assumed to provide local neural network weights, which are modeled through our framework. We then develop an inference approach that allows us to synthesize a more expressive global network without additional supervision, data pooling and with as few as a single communication round. We then demonstrate the efficacy of our approach on federated learning problems simulated from two popular image classification datasets.
We introduce a new second-order inertial optimization method for machine learning called INNA. It exploits the geometry of the loss function while only requiring stochastic approximations of the function values and the generalized gradients. This makes INNA fully implementable and adapted to large-scale optimization problems such as the training of deep neural networks. The algorithm combines both gradient-descent and Newton-like behaviors as well as inertia. We prove the convergence of INNA for most deep learning problems. To do so, we provide a well-suited framework to analyze deep learning loss functions involving tame optimization in which we study a continuous dynamical system together with its discrete stochastic approximations. We prove sublinear convergence for the continuous-time differential inclusion which underlies our algorithm. Additionally, we also show how standard optimization mini-batch methods applied to non-smooth non-convex problems can yield a certain type of spurious stationary points never discussed before. We address this issue by providing a theoretical framework around the new idea of $D$-criticality; we then give a simple asymptotic analysis of INNA. Our algorithm allows for using an aggressive learning rate of $o(1/\log k)$. From an empirical viewpoint, we show that INNA returns competitive results with respect to state of the art (stochastic gradient descent, ADAGRAD, ADAM) on popular deep learning benchmark problems.
Electroencephalography (EEG) is a useful way to implicitly monitor the users perceptual state during multimedia consumption. One of the primary challenges for the practical use of EEG-based monitoring is to achieve a satisfactory level of accuracy in EEG classification. Connectivity between different brain regions is an important property for the classification of EEG. However, how to define the connectivity structure for a given task is still an open problem, because there is no ground truth about how the connectivity structure should be in order to maximize the classification performance. In this paper, we propose an end-to-end neural network model for EEG-based emotional video classification, which can extract an appropriate multi-layer graph structure and signal features directly from a set of raw EEG signals and perform classification using them. Experimental results demonstrate that our method yields improved performance in comparison to the existing approaches where manually defined connectivity structures and signal features are used. Furthermore, we show that the graph structure extraction process is reliable in terms of consistency, and the learned graph structures make much sense in the viewpoint of emotional perception occurring in the brain.
We propose a computationally efficient wrapper feature selection method - called Autoencoder and Model Based Elimination of features using Relevance and Redundancy scores (AMBER) - that uses a single ranker model along with autoencoders to perform greedy backward elimination of features. The ranker model is used to prioritize the removal of features that are not critical to the classification task, while the autoencoders are used to prioritize the elimination of correlated features. We demonstrate the superior feature selection ability of AMBER on 4 well known datasets corresponding to different domain applications via comparing the classification accuracies with other computationally efficient state-of-the-art feature selection techniques. Interestingly, we find that the ranker model that is used for feature selection does not necessarily have to be the same as the final classifier that is trained on the selected features. Finally, we note how a smaller number of features can lead to higher accuracies on some datasets, and hypothesize that overfitting the ranker model on the training set facilitates the selection of more salient features.
The vulnerabilities of deep neural networks against adversarial examples have become a significant concern for deploying these models in sensitive domains. Devising a definitive defense against such attacks is proven to be challenging, and the methods relying on detecting adversarial samples are only valid when the attacker is oblivious to the detection mechanism. In this paper we propose a principled adversarial example detection method that can withstand norm-constrained white-box attacks. Inspired by one-versus-the-rest classification, in a K class classification problem, we train K binary classifiers where the i-th binary classifier is used to distinguish between clean data of class i and adversarially perturbed samples of other classes. At test time, we first use a trained classifier to get the predicted label (say k) of the input, and then use the k-th binary classifier to determine whether the input is a clean sample (of class k) or an adversarially perturbed example (of other classes). We further devise a generative approach to detecting/classifying adversarial examples by interpreting each binary classifier as an unnormalized density model of the class-conditional data. We provide comprehensive evaluation of the above adversarial example detection/classification methods, and demonstrate their competitive performances and compelling properties.
In this work, we develop a novel regularizer to improve the learning of long-range dependency of sequence data. Applied on language modelling, our regularizer expresses the inductive bias that sequence variables should have high mutual information even though the model might not see abundant observations for complex long-range dependency. We show how the `next sentence prediction (classification)' heuristic can be derived in a principled way from our mutual information estimation framework, and be further extended to maximize the mutual information of sequence variables. The proposed approach not only is effective at increasing the mutual information of segments under the learned model but more importantly, leads to a higher likelihood on holdout data, and improved generation quality. Code is released at https://github.com/BorealisAI/BMI.
Music Genre Classification is the problem of associating genre-related labels to digitized music tracks. It has applications in the organization of commercial and personal music collections. Often, music tracks are described as a set of timbre-inspired sound textures. In shallow-learning systems, the total number of sound textures per track is usually too high, and texture downsampling is necessary to make training tractable. Although previous work has solved this by linear downsampling, no extensive work has been done to evaluate how texture selection benefits genre classification in the context of the bag of frames track descriptions. In this paper, we evaluate the impact of frame selection on automatic music genre classification in a bag of frames scenario. We also present a novel texture selector based on K-Means aimed to identify diverse sound textures within each track. We evaluated texture selection in diverse datasets, four different feature sets, as well as its relationship to a univariate feature selection strategy. The results show that frame selection leads to significant improvement over the single vector baseline on datasets consisting of full-length tracks, regardless of the feature set. Results also indicate that the K-Means texture selector achieves significant improvements over the baseline, using fewer textures per track than the commonly used linear downsampling. The results also suggest that texture selection is complementary to the feature selection strategy evaluated. Our qualitative analysis indicates that texture variety within classes benefits model generalization. Our analysis shows that selecting specific audio excerpts can improve classification performance, and it can be done automatically.
We provide a theoretical explanation for the effectiveness of gradient clipping in training deep neural networks. The key ingredient is a new smoothness condition derived from practical neural network training examples. We observe that gradient smoothness, a concept central to the analysis of first-order optimization algorithms that is often assumed to be a constant, demonstrates significant variability along the training trajectory of deep neural networks. Further, this smoothness positively correlates with the gradient norm, and contrary to standard assumptions in the literature, it can grow with the norm of the gradient. These empirical observations limit the applicability of existing theoretical analyses of algorithms that rely on a fixed bound on smoothness. These observations motivate us to introduce a novel relaxation of gradient smoothness that is weaker than the commonly used Lipschitz smoothness assumption. Under the new condition, we prove that two popular methods, namely, \emph{gradient clipping} and \emph{normalized gradient}, converge arbitrarily faster than gradient descent with fixed stepsize. We further explain why such adaptively scaled gradient methods can accelerate empirical convergence and verify our results empirically in popular neural network training settings.
We investigate adversarial robustness of Gaussian Process Classification (GPC) models. Given a compact subset of the input space $T\subseteq \mathbb{R}^d$ enclosing a test point $x^*$ and a GPC trained on a dataset $\mathcal{D}$, we aim to compute the minimum and the maximum classification probability for the GPC over all the points in $T$. In order to do so, we show how functions lower- and upper-bounding the GPC output in $T$ can be derived, and implement those in a branch and bound optimisation algorithm. For any error threshold $\epsilon > 0$ selected a priori, we show that our algorithm is guaranteed to reach values $\epsilon$-close to the actual values in finitely many iterations. We apply our method to investigate the robustness of GPC models on a 2D synthetic dataset, the SPAM dataset and a subset of the MNIST dataset, providing comparisons of different GPC training techniques, and show how our method can be used for interpretability analysis. Our empirical analysis suggests that GPC robustness increases with more accurate posterior estimation.