Research connected to &quot;machine learning&quot;

Feature-weighted Stacking for Nonseasonal Time Series Forecasts: A Case Study of the COVID-19 Epidemic Curves

We investigate ensembling techniques in forecasting and examine their potential for use in nonseasonal time-series similar to those in the early days of the COVID-19 pandemic. Developing improved forecast methods is essential as they provide data-driven decisions to organisations and decision-makers during critical phases. We propose using late data fusion, using a stacked ensemble of two forecasting models and two meta-features that prove their predictive power during a preliminary forecasting stage. The final ensembles include a Prophet and long short term memory (LSTM) neural network as base models. The base models are combined by a multilayer perceptron (MLP), taking into account meta-features that indicate the highest correlation with each base model's forecast accuracy. We further show that the inclusion of meta-features generally improves the ensemble's forecast accuracy across two forecast horizons of seven and fourteen days. This research reinforces previous work and demonstrates the value of combining traditional statistical models with deep learning models to produce more accurate forecast models for time-series from different domains and seasonality.

preprint2015arXiv

An Experimental Comparison of Several Clustering and Initialization Methods

We examine methods for clustering in high dimensions. In the first part of the paper, we perform an experimental comparison between three batch clustering algorithms: the Expectation-Maximization (EM) algorithm, a winner take all version of the EM algorithm reminiscent of the K-means algorithm, and model-based hierarchical agglomerative clustering. We learn naive-Bayes models with a hidden root node, using high-dimensional discrete-variable data sets (both real and synthetic). We find that the EM algorithm significantly outperforms the other methods, and proceed to investigate the effect of various initialization schemes on the final solution produced by the EM algorithm. The initializations that we consider are (1) parameters sampled from an uninformative prior, (2) random perturbations of the marginal distribution of the data, and (3) the output of hierarchical agglomerative clustering. Although the methods are substantially different, they lead to learned models that are strikingly similar in quality.

LotteryFL: Personalized and Communication-Efficient Federated Learning with Lottery Ticket Hypothesis on Non-IID Datasets

Federated learning is a popular distributed machine learning paradigm with enhanced privacy. Its primary goal is learning a global model that offers good performance for the participants as many as possible. The technology is rapidly advancing with many unsolved challenges, among which statistical heterogeneity (i.e., non-IID) and communication efficiency are two critical ones that hinder the development of federated learning. In this work, we propose LotteryFL -- a personalized and communication-efficient federated learning framework via exploiting the Lottery Ticket hypothesis. In LotteryFL, each client learns a lottery ticket network (i.e., a subnetwork of the base model) by applying the Lottery Ticket hypothesis, and only these lottery networks will be communicated between the server and clients. Rather than learning a shared global model in classic federated learning, each client learns a personalized model via LotteryFL; the communication cost can be significantly reduced due to the compact size of lottery networks. To support the training and evaluation of our framework, we construct non-IID datasets based on MNIST, CIFAR-10 and EMNIST by taking feature distribution skew, la

Online Learning-to-Defer with Varying Experts

Learning-to-Defer (L2D) methods route each query either to a predictive model or to external experts. While existing work studies this problem in batch settings, real-world deployments require handling streaming data, changing expert availability, and shifting expert distribution. We introduce the first online L2D algorithm for multiclass classification with bandit feedback and a dynamically varying pool of experts. Our method achieves regret guarantees of $O((n+n_e)T^{2/3})$ in general and $O((n+n_e)\sqrt{T})$ under a low-noise condition, where $T$ is the time horizon, $n$ is the number of labels, and $n_e$ is the number of distinct experts observed across rounds. The analysis builds on novel $\mathcal{H}$-consistency bounds for the online framework, combined with first-order methods for online convex optimization. Experiments on synthetic and real-world datasets demonstrate that our approach effectively extends standard Learning-to-Defer to settings with varying expert availability and reliability.

Cascading Hybrid Bandits: Online Learning to Rank for Relevance and Diversity

Relevance ranking and result diversification are two core areas in modern recommender systems. Relevance ranking aims at building a ranked list sorted in decreasing order of item relevance, while result diversification focuses on generating a ranked list of items that covers a broad range of topics. In this paper, we study an online learning setting that aims to recommend a ranked list with $K$ items that maximizes the ranking utility, i.e., a list whose items are relevant and whose topics are diverse. We formulate it as the cascade hybrid bandits (CHB) problem. CHB assumes the cascading user behavior, where a user browses the displayed list from top to bottom, clicks the first attractive item, and stops browsing the rest. We propose a hybrid contextual bandit approach, called CascadeHybrid, for solving this problem. CascadeHybrid models item relevance and topical diversity using two independent functions and simultaneously learns those functions from user click feedback. We conduct experiments to evaluate CascadeHybrid on two real-world recommendation datasets: MovieLens and Yahoo music datasets. Our experimental results show that CascadeHybrid outperforms the baselines. In additi

preprint2015arXiv

A distributed block coordinate descent method for training $l_1$ regularized linear classifiers

Distributed training of $l_1$ regularized classifiers has received great attention recently. Most existing methods approach this problem by taking steps obtained from approximating the objective by a quadratic approximation that is decoupled at the individual variable level. These methods are designed for multicore and MPI platforms where communication costs are low. They are inefficient on systems such as Hadoop running on a cluster of commodity machines where communication costs are substantial. In this paper we design a distributed algorithm for $l_1$ regularization that is much better suited for such systems than existing algorithms. A careful cost analysis is used to support these points and motivate our method. The main idea of our algorithm is to do block optimization of many variables on the actual objective function within each computing node; this increases the computational cost per step that is matched with the communication cost, and decreases the number of outer iterations, thus yielding a faster overall method. Distributed Gauss-Seidel and Gauss-Southwell greedy schemes are used for choosing variables to update in each step. We establish global convergence theory for

TPG-DNN: A Method for User Intent Prediction Based on Total Probability Formula and GRU Loss with Multi-task Learning

The E-commerce platform has become the principal battleground where people search, browse and pay for whatever they want. Critical as is to improve the online shopping experience for customers and merchants, how to find a proper approach for user intent prediction are paid great attention in both industry and academia. In this paper, we propose a novel user intent prediction model, TPG-DNN, to complete the challenging task, which is based on adaptive gated recurrent unit (GRU) loss function with multi-task learning. We creatively use the GRU structure and total probability formula as the loss function to model the users' whole online purchase process. Besides, the multi-task weight adjustment mechanism can make the final loss function dynamically adjust the importance between different tasks through data variance. According to the test result of experiments conducted on Taobao daily and promotion data sets, the proposed model performs much better than existing click through rate (CTR) models. At present, the proposed user intent prediction model has been widely used for the coupon allocation, advertisement and recommendation on Taobao platform, which greatly improve the user ex

Asymptotic Accuracy of Distribution-Based Estimation for Latent Variables

Hierarchical statistical models are widely employed in information science and data engineering. The models consist of two types of variables: observable variables that represent the given data and latent variables for the unobservable labels. An asymptotic analysis of the models plays an important role in evaluating the learning process; the result of the analysis is applied not only to theoretical but also to practical situations, such as optimal model selection and active learning. There are many studies of generalization errors, which measure the prediction accuracy of the observable variables. However, the accuracy of estimating the latent variables has not yet been elucidated. For a quantitative evaluation of this, the present paper formulates distribution-based functions for the errors in the estimation of the latent variables. The asymptotic behavior is analyzed for both the maximum likelihood and the Bayes methods.

On the computational and statistical complexity of over-parameterized matrix sensing

We consider solving the low rank matrix sensing problem with Factorized Gradient Descend (FGD) method when the true rank is unknown and over-specified, which we refer to as over-parameterized matrix sensing. If the ground truth signal $\mathbf{X}^* \in \mathbb{R}^{d*d}$ is of rank $r$, but we try to recover it using $\mathbf{F} \mathbf{F}^\top$ where $\mathbf{F} \in \mathbb{R}^{d*k}$ and $k>r$, the existing statistical analysis falls short, due to a flat local curvature of the loss function around the global maxima. By decomposing the factorized matrix $\mathbf{F}$ into separate column spaces to capture the effect of extra ranks, we show that $\|\mathbf{F}_t \mathbf{F}_t - \mathbf{X}^*\|_{F}^2$ converges to a statistical error of $\tilde{\mathcal{O}} ({k d σ^2/n})$ after $\tilde{\mathcal{O}}(\frac{σ_{r}}σ\sqrt{\frac{n}{d}})$ number of iterations where $\mathbf{F}_t$ is the output of FGD after $t$ iterations, $σ^2$ is the variance of the observation noise, $σ_{r}$ is the $r$-th largest eigenvalue of $\mathbf{X}^*$, and $n$ is the number of sample. Our results, therefore, offer a comprehensive picture of the statistical and computational complexity of FGD for the over-parameterized matrix sensing problem.

Estimating nonlinear regression errors without doing regression

A method for estimating nonlinear regression errors and their distributions without performing regression is presented. Assuming continuity of the modeling function the variance is given in terms of conditional probabilities extracted from the data. For N data points the computational demand is N2. Comparing the predicted residual errors with those derived from a linear model assumption provides a signal for nonlinearity. The method is successfully illustrated with data generated by the Ikeda and Lorenz maps augmented with noise. As a by-product the embedding dimensions of these maps are also extracted.

Information Plane Analysis of Binary Neural Networks

Information plane (IP) analysis has been suggested to study the training dynamics of deep neural networks through mutual information (MI) between inputs, representations, and targets. However, its statistical validity is often compromised by the difficulty of estimating MI from samples of high-dimensional, deterministic representations. In this work, we perform IP analyses on binary neural networks (BNNs) where activations are discrete and MI is finite. We characterise the finite-sample behaviour of the plug-in entropy estimator and identify regimes for sample size $N$ and representation dimensionality $D$ under which MI estimates are reliable. Outside these regimes, we show that empirical MI estimates saturate to $\log_2 N$, rendering IP trajectories uninformative. Restricting attention to the reliable regime, we train 375 BNNs to investigate the existence of late-stage compression phases and the relationship between compressed representations and generalisation performance. Our results show that while late-stage compression is frequently observed, compressed latent representations do not consistently correlate with improved generalization performance. Instead, the relationship between compression and generalisation is highly dependent on task, architecture, and regularisation.

CEPO: RLVR Self-Distillation using Contrastive Evidence Policy Optimization

When a model produces a correct solution under reinforcement learning with verifiable rewards (RLVR), every token receives the same reward signal regardless of whether it was a decisive reasoning step or a grammatical filler. A natural fix is to condition the model on the correct answer as a teacher, identifying tokens it would have generated differently had it known the answer. Prior work shows this either corrupts training by leaking the answer into the gradient, or produces a weak signal that cannot distinguish decisive steps from filler, since both look equally surprising relative to the model's baseline. We propose Contrastive Evidence Policy Optimization (CEPO), which asks a sharper question at every token: not just "does the correct answer favor this token?" but "does the correct answer favor it while the wrong answer disfavors it?" A token satisfying both is a genuine reasoning step; one satisfying neither is filler. The wrong-answer teacher is constructed from rejected rollouts already in the training batch, incurring no additional sampling cost. We prove CEPO inherits all structural safety guarantees of the prior state of the art while strictly sharpening credit at decisive tokens, with the improvement vanishing exactly at filler positions. Empirically, CEPO achieves 43.43% and 60.56% average accuracy across five multimodal mathematical reasoning benchmarks at 2B and 4B scale, respectively, versus 41.17% and 57.43% for GRPO under identical training budgets. Distribution-matching self-distillation methods (OPSD, SDPO) fall below the untrained baseline, empirically confirming the information leakage our theory predicts. Our code is available at https://github.com/ahmedheakl/CEPO.

NMF with Sparse Regularizations in Transformed Domains

Non-negative blind source separation (non-negative BSS), which is also referred to as non-negative matrix factorization (NMF), is a very active field in domains as different as astrophysics, audio processing or biomedical signal processing. In this context, the efficient retrieval of the sources requires the use of signal priors such as sparsity. If NMF has now been well studied with sparse constraints in the direct domain, only very few algorithms can encompass non-negativity together with sparsity in a transformed domain since simultaneously dealing with two priors in two different domains is challenging. In this article, we show how a sparse NMF algorithm coined non-negative generalized morphological component analysis (nGMCA) can be extended to impose non-negativity in the direct domain along with sparsity in a transformed domain, with both analysis and synthesis formulations. To our knowledge, this work presents the first comparison of analysis and synthesis priors ---as well as their reweighted versions--- in the context of blind source separation. Comparisons with state-of-the-art NMF algorithms on realistic data show the efficiency as well as the robustness of the proposed

VAE Approximation Error: ELBO and Exponential Families

The importance of Variational Autoencoders reaches far beyond standalone generative models -- the approach is also used for learning latent representations and can be generalized to semi-supervised learning. This requires a thorough analysis of their commonly known shortcomings: posterior collapse and approximation errors. This paper analyzes VAE approximation errors caused by the combination of the ELBO objective and encoder models from conditional exponential families, including, but not limited to, commonly used conditionally independent discrete and continuous models. We characterize subclasses of generative models consistent with these encoder families. We show that the ELBO optimizer is pulled away from the likelihood optimizer towards the consistent subset and study this effect experimentally. Importantly, this subset can not be enlarged, and the respective error cannot be decreased, by considering deeper encoder/decoder networks.

Deep Learning for Moving Blockage Prediction using Real Millimeter Wave Measurements

Millimeter wave (mmWave) communication is a key component of 5G and beyond. Harvesting the gains of the large bandwidth and low latency at mmWave systems, however, is challenged by the sensitivity of mmWave signals to blockages; a sudden blockage in the line of sight (LOS) link leads to abrupt disconnection, which affects the reliability of the network. In addition, searching for an alternative base station to re-establish the link could result in needless latency overhead. In this paper, we address these challenges collectively by utilizing machine learning to anticipate dynamic blockages proactively. The proposed approach sees a machine learning algorithm learning to predict future blockages by observing what we refer to as the pre-blockage signature. To evaluate our proposed approach, we build a mmWave communication setup with a moving blockage and collect a dataset of received power sequences. Simulation results on a real dataset show that blockage occurrence could be predicted with more than 85% accuracy and the exact time instance of blockage occurrence can be obtained with low error. This highlights the potential of the proposed solution for dynamic blockage prediction and proactive hand-off, which enhances the reliability and latency of future wireless networks.

On Agnostic PAC Learning using $\mathcal{L}_2$-polynomial Regression and Fourier-based Algorithms

We develop a framework using Hilbert spaces as a proxy to analyze PAC learning problems with structural properties. We consider a joint Hilbert space incorporating the relation between the true label and the predictor under a joint distribution $D$. We demonstrate that agnostic PAC learning with 0-1 loss is equivalent to an optimization in the Hilbert space domain. With our model, we revisit the PAC learning problem using methods based on least-squares such as $\mathcal{L}_2$ polynomial regression and Linial's low-degree algorithm. We study learning with respect to several hypothesis classes such as half-spaces and polynomial-approximated classes (i.e., functions approximated by a fixed-degree polynomial). We prove that (under some distributional assumptions) such methods obtain generalization error up to $2opt$ with $opt$ being the optimal error of the class. Hence, we show the tightest bound on generalization error when $opt\leq 0.2$.

Factorial Hidden Markov Models for Learning Representations of Natural Language

Most representation learning algorithms for language and image processing are local, in that they identify features for a data point based on surrounding points. Yet in language processing, the correct meaning of a word often depends on its global context. As a step toward incorporating global context into representation learning, we develop a representation learning algorithm that incorporates joint prediction into its technique for producing features for a word. We develop efficient variational methods for learning Factorial Hidden Markov Models from large texts, and use variational distributions to produce features for each word that are sensitive to the entire input sequence, not just to a local context window. Experiments on part-of-speech tagging and chunking indicate that the features are competitive with or better than existing state-of-the-art representation learning methods.

preprint2015arXiv

Super-Resolution Off the Grid

Super-resolution is the problem of recovering a superposition of point sources using bandlimited measurements, which may be corrupted with noise. This signal processing problem arises in numerous imaging problems, ranging from astronomy to biology to spectroscopy, where it is common to take (coarse) Fourier measurements of an object. Of particular interest is in obtaining estimation procedures which are robust to noise, with the following desirable statistical and computational properties: we seek to use coarse Fourier measurements (bounded by some cutoff frequency); we hope to take a (quantifiably) small number of measurements; we desire our algorithm to run quickly. Suppose we have k point sources in d dimensions, where the points are separated by at least Δfrom each other (in Euclidean distance). This work provides an algorithm with the following favorable guarantees: - The algorithm uses Fourier measurements, whose frequencies are bounded by O(1/Δ) (up to log factors). Previous algorithms require a cutoff frequency which may be as large as Ω( d/Δ). - The number of measurements taken by and the computational complexity of our algorithm are bounded by a polynomial in both the num

Clinical outcome prediction under hypothetical interventions -- a representation learning framework for counterfactual reasoning

Most machine learning (ML) models are developed for prediction only; offering no option for causal interpretation of their predictions or parameters/properties. This can hamper the health systems' ability to employ ML models in clinical decision-making processes, where the need and desire for predicting outcomes under hypothetical investigations (i.e., counterfactual reasoning/explanation) is high. In this research, we introduce a new representation learning framework (i.e., partial concept bottleneck), which considers the provision of counterfactual explanations as an embedded property of the risk model. Despite architectural changes necessary for jointly optimising for prediction accuracy and counterfactual reasoning, the accuracy of our approach is comparable to prediction-only models. Our results suggest that our proposed framework has the potential to help researchers and clinicians improve personalised care (e.g., by investigating the hypothetical differential effects of interventions)

Norm-based generalisation bounds for multi-class convolutional neural networks

We show generalisation error bounds for deep learning with two main improvements over the state of the art. (1) Our bounds have no explicit dependence on the number of classes except for logarithmic factors. This holds even when formulating the bounds in terms of the $L^2$-norm of the weight matrices, where previous bounds exhibit at least a square-root dependence on the number of classes. (2) We adapt the classic Rademacher analysis of DNNs to incorporate weight sharing -- a task of fundamental theoretical importance which was previously attempted only under very restrictive assumptions. In our results, each convolutional filter contributes only once to the bound, regardless of how many times it is applied. Further improvements exploiting pooling and sparse connections are provided. The presented bounds scale as the norms of the parameter matrices, rather than the number of parameters. In particular, contrary to bounds based on parameter counting, they are asymptotically tight (up to log factors) when the weights approach initialisation, making them suitable as a basic ingredient in bounds sensitive to the optimisation procedure. We also show how to adapt the recent technique of loss function augmentation to our situation to replace spectral norms by empirical analogues whilst maintaining the advantages of our approach.

preprint2016arXiv

Extrapolation and learning equations

In classical machine learning, regression is treated as a black box process of identifying a suitable function from a hypothesis set without attempting to gain insight into the mechanism connecting inputs and outputs. In the natural sciences, however, finding an interpretable function for a phenomenon is the prime goal as it allows to understand and generalize results. This paper proposes a novel type of function learning network, called equation learner (EQL), that can learn analytical expressions and is able to extrapolate to unseen domains. It is implemented as an end-to-end differentiable feed-forward network and allows for efficient gradient based training. Due to sparsity regularization concise interpretable expressions can be obtained. Often the true underlying source expression is identified.

preprint2012arXiv

Learning AMP Chain Graphs under Faithfulness

This paper deals with chain graphs under the alternative Andersson-Madigan-Perlman (AMP) interpretation. In particular, we present a constraint based algorithm for learning an AMP chain graph a given probability distribution is faithful to. We also show that the extension of Meek's conjecture to AMP chain graphs does not hold, which compromises the development of efficient and correct score+search learning algorithms under assumptions weaker than faithfulness.

preprint2016arXiv

Recurrent Neural Network Training with Dark Knowledge Transfer

Recurrent neural networks (RNNs), particularly long short-term memory (LSTM), have gained much attention in automatic speech recognition (ASR). Although some successful stories have been reported, training RNNs remains highly challenging, especially with limited training data. Recent research found that a well-trained model can be used as a teacher to train other child models, by using the predictions generated by the teacher model as supervision. This knowledge transfer learning has been employed to train simple neural nets with a complex one, so that the final performance can reach a level that is infeasible to obtain by regular training. In this paper, we employ the knowledge transfer learning approach to train RNNs (precisely LSTM) using a deep neural network (DNN) model as the teacher. This is different from most of the existing research on knowledge transfer learning, since the teacher (DNN) is assumed to be weaker than the child (RNN); however, our experiments on an ASR task showed that it works fairly well: without applying any tricks on the learning scheme, this approach can train RNNs successfully even with limited training data.

Shape-CD: Change-Point Detection in Time-Series Data with Shapes and Neurons

Change-point detection in a time series aims to discover the time points at which some unknown underlying physical process that generates the time-series data has changed. We found that existing approaches become less accurate when the underlying process is complex and generates large varieties of patterns in the time series. To address this shortcoming, we propose Shape-CD, a simple, fast, and accurate change point detection method. Shape-CD uses shape-based features to model the patterns and a conditional neural field to model the temporal correlations among the time regions. We evaluated the performance of Shape-CD using four highly dynamic time-series datasets, including the ExtraSensory dataset with up to 2000 classes. Shape-CD demonstrated improved accuracy (7-60% higher in AUC) and faster computational speed compared to existing approaches. Furthermore, the Shape-CD model consists of only hundreds of parameters and require less data to train than other deep supervised learning models.

Efficient Chemical Space Exploration Using Active Learning Based on Marginalized Graph Kernel: an Application for Predicting the Thermodynamic Properties of Alkanes with Molecular Simulation

We introduce an explorative active learning (AL) algorithm based on Gaussian process regression and marginalized graph kernel (GPR-MGK) to explore chemical space with minimum cost. Using high-throughput molecular dynamics simulation to generate data and graph neural network (GNN) to predict, we constructed an active learning molecular simulation framework for thermodynamic property prediction. In specific, targeting 251,728 alkane molecules consisting of 4 to 19 carbon atoms and their liquid physical properties: densities, heat capacities, and vaporization enthalpies, we use the AL algorithm to select the most informative molecules to represent the chemical space. Validation of computational and experimental test sets shows that only 313 (0.124\% of the total) molecules were sufficient to train an accurate GNN model with $\rm R^2 > 0.99$ for computational test sets and $\rm R^2 > 0.94$ for experimental test sets. We highlight two advantages of the presented AL algorithm: compatibility with high-throughput data generation and reliable uncertainty quantification.

An End-to-End Graph Convolutional Kernel Support Vector Machine

A novel kernel-based support vector machine (SVM) for graph classification is proposed. The SVM feature space mapping consists of a sequence of graph convolutional layers, which generates a vector space representation for each vertex, followed by a pooling layer which generates a reproducing kernel Hilbert space (RKHS) representation for the graph. The use of a RKHS offers the ability to implicitly operate in this space using a kernel function without the computational complexity of explicitly mapping into it. The proposed model is trained in a supervised end-to-end manner whereby the convolutional layers, the kernel function and SVM parameters are jointly optimized with respect to a regularized classification loss. This approach is distinct from existing kernel-based graph classification models which instead either use feature engineering or unsupervised learning to define the kernel function. Experimental results demonstrate that the proposed model outperforms existing deep learning baseline models on a number of datasets.

syslrn: Learning What to Monitor for Efficient Anomaly Detection

While monitoring system behavior to detect anomalies and failures is important, existing methods based on log-analysis can only be as good as the information contained in the logs, and other approaches that look at the OS-level software state introduce high overheads. We tackle the problem with syslrn, a system that first builds an understanding of a target system offline, and then tailors the online monitoring instrumentation based on the learned identifiers of normal behavior. While our syslrn prototype is still preliminary and lacks many features, we show in a case study for the monitoring of OpenStack failures that it can outperform state-of-the-art log-analysis systems with little overhead.

preprint2012arXiv

PAC-Bayes-Bernstein Inequality for Martingales and its Application to Multiarmed Bandits

We develop a new tool for data-dependent analysis of the exploration-exploitation trade-off in learning under limited feedback. Our tool is based on two main ingredients. The first ingredient is a new concentration inequality that makes it possible to control the concentration of weighted averages of multiple (possibly uncountably many) simultaneously evolving and interdependent martingales. The second ingredient is an application of this inequality to the exploration-exploitation trade-off via importance weighted sampling. We apply the new tool to the stochastic multiarmed bandit problem, however, the main importance of this paper is the development and understanding of the new tool rather than improvement of existing algorithms for stochastic multiarmed bandits. In the follow-up work we demonstrate that the new tool can improve over state-of-the-art in structurally richer problems, such as stochastic multiarmed bandits with side information (Seldin et al., 2011a).

Defending against substitute model black box adversarial attacks with the 01 loss

Substitute model black box attacks can create adversarial examples for a target model just by accessing its output labels. This poses a major challenge to machine learning models in practice, particularly in security sensitive applications. The 01 loss model is known to be more robust to outliers and noise than convex models that are typically used in practice. Motivated by these properties we present 01 loss linear and 01 loss dual layer neural network models as a defense against transfer based substitute model black box attacks. We compare the accuracy of adversarial examples from substitute model black box attacks targeting our 01 loss models and their convex counterparts for binary classification on popular image benchmarks. Our 01 loss dual layer neural network has an adversarial accuracy of 66.2%, 58%, 60.5%, and 57% on MNIST, CIFAR10, STL10, and ImageNet respectively whereas the sigmoid activated logistic loss counterpart has accuracies of 63.5%, 19.3%, 14.9%, and 27.6%. Except for MNIST the convex counterparts have substantially lower adversarial accuracies. We show practical applications of our models to deter traffic sign and facial recognition adversarial attacks. On GTS

preprint2016arXiv

Unsupervised learning of object semantic parts from internal states of CNNs by population encoding

We address the key question of how object part representations can be found from the internal states of CNNs that are trained for high-level tasks, such as object classification. This work provides a new unsupervised method to learn semantic parts and gives new understanding of the internal representations of CNNs. Our technique is based on the hypothesis that semantic parts are represented by populations of neurons rather than by single filters. We propose a clustering technique to extract part representations, which we call Visual Concepts. We show that visual concepts are semantically coherent in that they represent semantic parts, and visually coherent in that corresponding image patches appear very similar. Also, visual concepts provide full spatial coverage of the parts of an object, rather than a few sparse parts as is typically found in keypoint annotations. Furthermore, We treat single visual concept as part detector and evaluate it for keypoint detection using the PASCAL3D+ dataset and for part detection using our newly annotated ImageNetPart dataset. The experiments demonstrate that visual concepts can be used to detect parts. We also show that some visual concepts respo

Hidden Markov Models with Momentum

Momentum is a popular technique for improving convergence rates during gradient descent. In this research, we experiment with adding momentum to the Baum-Welch expectation-maximization algorithm for training Hidden Markov Models. We compare discrete Hidden Markov Models trained with and without momentum on English text and malware opcode data. The effectiveness of momentum is determined by measuring the changes in model score and classification accuracy due to momentum. Our extensive experiments indicate that adding momentum to Baum-Welch can reduce the number of iterations required for initial convergence during HMM training, particularly in cases where the model is slow to converge. However, momentum does not seem to improve the final model performance at a high number of iterations.