Assessing the predictive accuracy of black box classifiers is challenging in the absence of labeled test datasets. In these scenarios we may need to rely on a human oracle to evaluate individual predictions; presenting the challenge to create query algorithms to guide the search for points that provide the most information about the classifier's predictive characteristics. Previous works have focused on developing utility models and query algorithms for discovering unknown unknowns --- misclassifications with a predictive confidence above some arbitrary threshold. However, if misclassifications occur at the rate reflected by the confidence values, then these search methods reveal nothing more than a proper assessment of predictive certainty. We are unable to properly mitigate the risks associated with model deficiency when the model's confidence in prediction exceeds the actual model accuracy. We propose a facility locations utility model and corresponding greedy query algorithm that instead searches for overconfident unknown unknowns. Through robust empirical experiments we demonstrate that the greedy query algorithm with the facility locations utility model consistently results in oracle queries with superior performance in discovering overconfident unknown unknowns than previous methods.

The challenge asks to provide a prediction for (i) a destination and the (ii) arrival time of ships in a streaming-fashion using Geo-spatial data in the maritime context. Novel aspects of our approach include the use of ensemble learning based on Random Forest, Gradient Boosting Decision Trees (GBDT), XGBoost Trees and Extremely Randomized Trees (ERT) in order to provide a prediction for a destination while for the arrival time, we propose the use of Feed-forward Neural Networks. In our evaluation, we were able to achieve an accuracy of 97% for the port destination classification problem and 90% (in mins) for the ETA prediction.

Many probabilistic models of interest in scientific computing and machine learning have expensive, black-box likelihoods that prevent the application of standard techniques for Bayesian inference, such as MCMC, which would require access to the gradient or a large number of likelihood evaluations. We introduce here a novel sample-efficient inference framework, Variational Bayesian Monte Carlo (VBMC). VBMC combines variational inference with Gaussian-process based, active-sampling Bayesian quadrature, using the latter to efficiently approximate the intractable integral in the variational objective. Our method produces both a nonparametric approximation of the posterior distribution and an approximate lower bound of the model evidence, useful for model selection. We demonstrate VBMC both on several synthetic likelihoods and on a neuronal model with data from real neurons. Across all tested problems and dimensions (up to $D = 10$), VBMC performs consistently well in reconstructing the posterior and the model evidence with a limited budget of likelihood evaluations, unlike other methods that work only in very low dimensions. Our framework shows great promise as a novel tool for posterior and model inference with expensive, black-box likelihoods.

High-risk domains require reliable confidence estimates from predictive models. Deep latent variable models provide these, but suffer from the rigid variational distributions used for tractable inference, which err on the side of overconfidence. We propose Stochastic Quantized Activation Distributions (SQUAD), which imposes a flexible yet tractable distribution over discretized latent variables. The proposed method is scalable, self-normalizing and sample efficient. We demonstrate that the model fully utilizes the flexible distribution, learns interesting non-linearities, and provides predictive uncertainty of competitive quality.

Popular machine learning estimators involve regularization parameters that can be challenging to tune, and standard strategies rely on grid search for this task. In this paper, we revisit the techniques of approximating the regularization path up to predefined tolerance $\epsilon$ in a unified framework and show that its complexity is $O(1/\sqrt[d]{\epsilon})$ for uniformly convex loss of order $d>0$ and $O(1/\sqrt{\epsilon})$ for Generalized Self-Concordant functions. This framework encompasses least-squares but also logistic regression (a case that as far as we know was not handled as precisely by previous works). We leverage our technique to provide refined bounds on the validation error as well as a practical algorithm for hyperparameter tuning. The later has global convergence guarantee when targeting a prescribed accuracy on the validation set. Last but not least, our approach helps relieving the practitioner from the (often neglected) task of selecting a stopping criterion when optimizing over the training set: our method automatically calibrates it based on the targeted accuracy on the validation set.

Shuffled linear regression is the problem of performing a linear regression fit to a dataset for which the correspondences between the independent samples and the observations are unknown. Such a problem arises in diverse domains such as computer vision, communications and biology. In its simplest form, it is tantamount to solving a linear system of equations, for which the entries of the right hand side vector have been permuted. This type of data corruption renders the linear regression task considerably harder, even in the absence of other corruptions, such as noise, outliers or missing entries. Existing methods are either applicable only to noiseless data or they are very sensitive to initialization and work only for partially shuffled data. In this paper we address both of these issues via an algebraic geometric approach, which uses symmetric polynomials to extract permutation-invariant constraints that the parameters $\boldsymbol{x} \in \mathbb{R}^n$ of the linear regression model must satisfy. This naturally leads to a polynomial system of $n$ equations in $n$ unknowns, which contains $\boldsymbol{x}$ in its root locus. Using the machinery of algebraic geometry we prove that as long as the independent samples are generic, this polynomial system is always consistent with at most $n!$ complex roots, regardless of any type of corruption inflicted on the observations. The algorithmic implication of this fact is that one can always solve this polynomial system and use its most suitable root as initialization to the Expectation Maximization algorithm. To the best of our knowledge, the resulting method is the first working solution for small values of $n$ able to handle thousands of fully shuffled noisy observations in milliseconds.

In this paper, we develop a non-parametric Bayesian estimation of Hawkes process kernel functions. Our method is based on the cluster representation of Hawkes processes. We sample random branching structures, and thus split the Hawkes process into clusters of Poisson processes, where the intensity function of each of these processes is the nonparametric triggering kernel of the Hawkes process. We derive both a block Gibbs sampler and a maximum a posteriori estimator based on stochastic expectation maximization. On synthetic data, we show our method to be flexible and scalable, and on two largescale Twitter diffusion datasets, we show our method to outperform the parametric Hawkes model. We observe that the learned non-parametric kernel reflects the longevity of different content types. Code has been made publicly available.

Making decisions requires information relevant to the task at hand. Many real-life decision-making situations allow acquiring further relevant information at a specific cost. For example, in assessing the health status of a patient we may decide to take additional measurements such as diagnostic tests or imaging scans before making a final assessment. More information that is relevant allows for better decisions but it may be costly to acquire all of this information. How can we trade off the desire to make good decisions with the option to acquire further information at a cost? To this end, we propose a principled framework, named EDDI (Efficient Dynamic Discovery of high-value Information), based on the theory of Bayesian experimental design. In EDDI we propose a novel partial variational autoencoder (Partial VAE), to efficiently handle missing data over varying subsets of known information. EDDI combines this Partial VAE with an acquisition function that maximizes expected information gain on a set of target variables. EDDI is efficient and demonstrates that dynamic discovery of high-value information is possible; we show cost reduction at the same decision quality and improved decision quality at the same cost in benchmarks and in two health-care applications. We believe there is great potential for realizing these gains in real-world decision support systems.

This paper presents a technology for simple and computationally efficient improvements of a generic Artificial Intelligence (AI) system, including Multilayer and Deep Learning neural networks. The improvements are, in essence, small network ensembles constructed on top of the existing AI architectures. Theoretical foundations of the technology are based on Stochastic Separation Theorems and the ideas of the concentration of measure. We show that, subject to mild technical assumptions on statistical properties of internal signals in the original AI system, the technology enables instantaneous and computationally efficient removal of spurious and systematic errors with probability close to one on the datasets which are exponentially large in dimension. The method is illustrated with numerical examples and a case study of ten digits recognition from American Sign Language.

The widespread use of GPS-enabled devices generates voluminous and continuous amounts of traffic data but analyzing such data for interpretable and actionable insights poses challenges. A hierarchical clustering of the trips has many uses such as discovering shortest paths, common routes and often traversed areas. However, hierarchical clustering typically has time complexity of $O(n^2 \log n)$ where $n$ is the number of instances, and is difficult to scale to large data sets associated with GPS data. Furthermore, incremental hierarchical clustering is still a developing area. Prefix trees (also called tries) can be efficiently constructed and updated in linear time (in $n$). We show how a specially constructed trie can compactly store the trips and further show this trie is equivalent to a dendrogram that would have been built by classic agglomerative hierarchical algorithms using a specific distance metric. This allows creating hierarchical clusterings of GPS trip data and updating this hierarchy in linear time. %we can extract a meaningful kernel and can also interpret the structure as clusterings of differing granularity as one progresses down the tree. We demonstrate the usefulness of our proposed approach on a real world data set of half a million taxis' GPS traces, well beyond the capabilities of agglomerative clustering methods. Our work is not limited to trip data and can be used with other data with a string representation.