Missing responses is a missing data format in which outcomes are not always observed. In this work we develop kernel machines that can handle missing responses. First, we propose a kernel machine family that uses mainly the complete cases. For the quadratic loss, we then propose a family of doubly-robust kernel machines. The proposed kernel-machine estimators can be applied to both regression and classification problems. We prove oracle inequalities for the finite-sample differences between the kernel machine risk and Bayes risk. We use these oracle inequalities to prove consistency and to calculate convergence rates. We demonstrate the performance of the two proposed kernel machine families using both a simulation study and a real-world data analysis.

Sentiment classification involves quantifying the affective reaction of a human to a document, media item or an event. Although researchers have investigated several methods to reliably infer sentiment from lexical, speech and body language cues, training a model with a small set of labeled datasets is still a challenge. For instance, in expanding sentiment analysis to new languages and cultures, it may not always be possible to obtain comprehensive labeled datasets. In this paper, we investigate the application of semi-supervised and transfer learning methods to improve performances on low resource sentiment classification tasks. We experiment with extracting dense feature representations, pre-training and manifold regularization in enhancing the performance of sentiment classification systems. Our goal is a coherent implementation of these methods and we evaluate the gains achieved by these methods in matched setting involving training and testing on a single corpus setting as well as two cross corpora settings. In both the cases, our experiments demonstrate that the proposed methods can significantly enhance the model performance against a purely supervised approach, particularly in cases involving a handful of training data.

Stochastic Gradient Langevin Dynamics (SGLD) is a sampling scheme for Bayesian modeling adapted to large datasets and models. SGLD relies on the injection of Gaussian Noise at each step of a Stochastic Gradient Descent (SGD) update. In this scheme, every component in the noise vector is independent and has the same scale, whereas the parameters we seek to estimate exhibit strong variations in scale and significant correlation structures, leading to poor convergence and mixing times. We compare different preconditioning approaches to the normalization of the noise vector and benchmark these approaches on the following criteria: 1) mixing times of the multivariate parameter vector, 2) regularizing effect on small dataset where it is easy to overfit, 3) covariate shift detection and 4) resistance to adversarial examples.

Randomized trials, also known as A/B tests, are used to select between two policies: a control and a treatment. Given a corresponding set of features, we can ideally learn an optimized policy P that maps the A/B test data features to action space and optimizes reward. However, although A/B testing provides an unbiased estimator for the value of deploying B (i.e., switching from policy A to B), direct application of those samples to learn the the optimized policy P generally does not provide an unbiased estimator of the value of P as the samples were observed when constructing P. In situations where the cost and risks associated of deploying a policy are high, such an unbiased estimator is highly desirable. We present a procedure for learning optimized policies and getting unbiased estimates for the value of deploying them. We wrap any policy learning procedure with a bagging process and obtain out-of-bag policy inclusion decisions for each sample. We then prove that inverse-propensity-weighting effect estimator is unbiased when applied to the optimized subset. Likewise, we apply the same idea to obtain out-of-bag unbiased per-sample value estimate of the measurement that is independent of the randomized treatment, and use these estimates to build an unbiased doubly-robust effect estimator. Lastly, we empirically shown that even when the average treatment effect is negative we can find a positive optimized policy.

Stein variational gradient decent (SVGD) has been shown to be a powerful approximate inference algorithm for complex distributions. However, the standard SVGD requires calculating the gradient of the target density and cannot be applied when the gradient is unavailable. In this work, we develop a gradient-free variant of SVGD (GF-SVGD), which replaces the true gradient with a surrogate gradient, and corrects the induced bias by re-weighting the gradients in a proper form. We show that our GF-SVGD can be viewed as the standard SVGD with a special choice of kernel, and hence directly inherits the theoretical properties of SVGD. We shed insights on the empirical choice of the surrogate gradient and propose an annealed GF-SVGD that leverages the idea of simulated annealing to improve the performance on high dimensional complex distributions. Empirical studies show that our method consistently outperforms a number of recent advanced gradient-free MCMC methods.

We introduce a new Bayesian multi-class support vector machine by formulating a pseudo-likelihood for a multi-class hinge loss in the form of a location-scale mixture of Gaussians. We derive a variational-inference-based training objective for gradient-based learning. Additionally, we employ an inducing point approximation which scales inference to large data sets. Furthermore, we develop hybrid Bayesian neural networks that combine standard deep learning components with the proposed model to enable learning for unstructured data. We provide empirical evidence that our model outperforms the competitor methods with respect to both training time and accuracy in classification experiments on 68 structured and two unstructured data sets. Finally, we highlight the key capability of our model in yielding prediction uncertainty for classification by demonstrating its effectiveness in the tasks of large-scale active learning and detection of adversarial images.

Recent studies have illustrated that stochastic gradient Markov Chain Monte Carlo techniques have a strong potential in non-convex optimization, where local and global convergence guarantees can be shown under certain conditions. By building up on this recent theory, in this study, we develop an asynchronous-parallel stochastic L-BFGS algorithm for non-convex optimization. The proposed algorithm is suitable for both distributed and shared-memory settings. We provide formal theoretical analysis and show that the proposed method achieves an ergodic convergence rate of ${\cal O}(1/\sqrt{N})$ ($N$ being the total number of iterations) and it can achieve a linear speedup under certain conditions. We perform several experiments on both synthetic and real datasets. The results support our theory and show that the proposed algorithm provides a significant speedup over the recently proposed synchronous distributed L-BFGS algorithm.

Datasets with significant proportions of noisy (incorrect) class labels present challenges for training accurate Deep Neural Networks (DNNs). We propose a new perspective for understanding DNN generalization for such datasets, by investigating the dimensionality of the deep representation subspace of training samples. We show that from a dimensionality perspective, DNNs exhibit quite distinctive learning styles when trained with clean labels versus when trained with a proportion of noisy labels. Based on this finding, we develop a new dimensionality-driven learning strategy, which monitors the dimensionality of subspaces during training and adapts the loss function accordingly. We empirically demonstrate that our approach is highly tolerant to significant proportions of noisy labels, and can effectively learn low-dimensional local subspaces that capture the data distribution.

Edwin V. Bonilla School of Computer Science and Engineering University of New South Wales Sydney, Australia We consider multi-task regression models where the observations are assumed to be a linear combination of several latent node functions and weight functions, which are both drawn from Gaussian process priors. Driven by the problem of developing scalable methods for distributed solar power forecasting, we propose coupled priors over groups of (node or weight) processes to estimate a forecast model for solar power production at multiple distributed sites, exploiting spatial dependence between functions. Our results show that our approach provides better quantification of predictive uncertainties than competing benchmarks while maintaining high point-prediction accuracy.

Identifying customer segments in retail banking portfolios with different risk profiles can improve the accuracy of credit scoring. The Variational Autoencoder (VAE) has shown promising results in different research domains, and it has been documented the powerful information embedded in the latent space of the VAE. We use the VAE and show that transforming the input data into a meaningful representation, it is possible to steer configurations in the latent space of the VAE. Specifically, the Weight of Evidence (WoE) transformation encapsulates the propensity to fall into financial distress and the latent space in the VAE preserves this characteristic in a well-defined clustering structure. These clusters have considerably different risk profiles and therefore are suitable not only for credit scoring but also for marketing and customer purposes. This new clustering methodology offers solutions to some of the challenges in the existing clustering algorithms, e.g., suggests the number of clusters, assigns cluster labels to new customers, enables cluster visualization, scales to large datasets, captures non-linear relationships among others. Finally, for portfolios with a large number of customers in each cluster, developing one classifier model per cluster can improve the credit scoring assessment.