Unsupervised Grammar Induction with Depth-bounded PCFG

There has been recent interest in applying cognitively or empirically motivated bounds on recursion depth to limit the search space of grammar induction models (Ponvert et al., 2011; Noji and Johnson, 2016; Shain et al., 2016). This work extends this depth-bounding approach to probabilistic context-free grammar induction (DB-PCFG), which has a smaller parameter space than hierarchical sequence models, and therefore more fully exploits the space reductions of depth-bounding. Results for this model on grammar acquisition from transcribed child-directed speech and newswire text exceed or are competitive with those of other models when evaluated on parse accuracy. Moreover, gram- mars acquired from this model demonstrate a consistent use of category labels, something which has not been demonstrated by other acquisition models.

How do Humans Understand Explanations from Machine Learning Systems? An Evaluation of the Human-Interpretability of Explanation

Recent years have seen a boom in interest in machine learning systems that can provide a human-understandable rationale for their predictions or decisions. However, exactly what kinds of explanation are truly human-interpretable remains poorly understood. This work advances our understanding of what makes explanations interpretable in the specific context of verification. Suppose we have a machine learning system that predicts X, and we provide rationale for this prediction X. Given an input, an explanation, and an output, is the output consistent with the input and the supposed rationale? Via a series of user-studies, we identify what kinds of increases in complexity have the greatest effect on the time it takes for humans to verify the rationale, and which seem relatively insensitive.

Prediction-Constrained Topic Models for Antidepressant Recommendation

Supervisory signals can help topic models discover low-dimensional data representations that are more interpretable for clinical tasks. We propose a framework for training supervised latent Dirichlet allocation that balances two goals: faithful generative explanations of high-dimensional data and accurate prediction of associated class labels. Existing approaches fail to balance these goals by not properly handling a fundamental asymmetry: the intended task is always predicting labels from data, not data from labels. Our new prediction-constrained objective trains models that predict labels from heldout data well while also producing good generative likelihoods and interpretable topic-word parameters. In a case study on predicting depression medications from electronic health records, we demonstrate improved recommendations compared to previous supervised topic models and high- dimensional logistic regression from words alone.

Improving the Adversarial Robustness and Interpretability of Deep Neural Networks by Regularizing their Input Gradients

Deep neural networks have proven remarkably effective at solving many classification problems, but have been criticized recently for two major weaknesses: the reasons behind their predictions are uninterpretable, and the predictions themselves can often be fooled by small adversarial perturbations. These problems pose major obstacles for the adoption of neural networks in domains that require security or transparency. In this work, we evaluate the effectiveness of defenses that differentiably penalize the degree to which small changes in inputs can alter model predictions. Across multiple attacks, architectures, defenses, and datasets, we find that neural networks trained with this input gradient regularization exhibit robustness to transferred adversarial examples generated to fool all of the other models. We also find that adversarial examples generated to fool gradient-regularized models fool all other models equally well, and actually lead to more "legitimate," interpretable misclassifications as rated by people (which we confirm in a human subject experiment). Finally, we demonstrate that regularizing input gradients makes them more naturally interpretable as rationales for model predictions. We conclude by discussing this relationship between interpretability and robustness in deep neural networks.

Accountability of AI Under the Law: The Role of Explanation

The ubiquity of systems using artificial intelligence or "AI" has brought increasing attention to how those systems should be regulated. The choice of how to regulate AI systems will require care. AI systems have the potential to synthesize large amounts of data, allowing for greater levels of personalization and precision than ever before---applications range from clinical decision support to autonomous driving and predictive policing. That said, there exist legitimate concerns about the intentional and unintentional negative consequences of AI systems. There are many ways to hold AI systems accountable. In this work, we focus on one: explanation. Questions about a legal right to explanation from AI systems was recently debated in the EU General Data Protection Regulation, and thus thinking carefully about when and how explanation from AI systems might improve accountability is timely. In this work, we review contexts in which explanation is currently required under the law, and then list the technical considerations that must be considered if we desired AI systems that could provide kinds of explanations that are currently required of humans.

Beyond Sparsity: Tree Regularization of Deep Models for Interpretability

The lack of interpretability remains a key barrier to the adoption of deep models in many applications. In this work, we explicitly regularize deep models so human users might step through the process behind their predictions in little time. Specifically, we train deep time-series models so their class-probability predictions have high accuracy while being closely modeled by decision trees with few nodes. Using intuitive toy examples as well as medical tasks for treating sepsis and HIV, we demonstrate that this new tree regularization yields models that are easier for humans to simulate than simpler L1 or L2 penalties without sacrificing predictive power.

Uncertainty Decomposition in Bayesian Neural Networks with Latent Variables

Bayesian neural networks (BNNs) with latent variables are probabilistic models which can automatically identify complex stochastic patterns in the data. We describe and study in these models a decomposition of predictive uncertainty into its epistemic and aleatoric components. First, we show how such a decomposition arises naturally in a Bayesian active learning scenario by following an information theoretic approach. Second, we use a similar decomposition to develop a novel risk sensitive objective for safe reinforcement learning (RL). This objective minimizes the effect of model bias in environments whose stochastic dynamics are described by BNNs with latent variables. Our experiments illustrate the usefulness of the resulting decomposition in active learning and safe RL settings.

Robust and Efficient Transfer Learning with Hidden-Parameter Markov Decision Processes

We introduce a new formulation of the Hidden Parameter Markov Decision Process (HiP-MDP), a framework for modeling families of related tasks using low-dimensional latent embeddings. Our new framework correctly models the joint uncertainty in the latent parameters and the state space. We also replace the original Gaussian Process-based model with a Bayesian Neural Network, enabling more scalable inference. Thus, we expand the scope of the HiP-MDP to applications with higher dimensions and more complex dynamics.

Weighted Tensor Decomposition for Learning Latent Variables with Partial Data

Tensor decomposition methods are popular tools for learning latent variables given only lower-order moments of the data. However, the standard assumption is that we have sufficient data to estimate these moments to high accuracy. In this work, we consider the case in which certain dimensions of the data are not always observed---common in applied settings, where not all measurements may be taken for all observations---resulting in moment estimates of varying quality. We derive a weighted tensor decomposition approach that is computationally as efficient as the non-weighted approach, and demonstrate that it outperforms methods that do not appropriately leverage these less-observed dimensions.

Prior matters: simple and general methods for evaluating and improving topic quality in topic modeling

Latent Dirichlet Allocation (LDA) models trained without stopword removal often produce topics with high posterior probabilities on uninformative words, obscuring the underlying corpus content. Even when canonical stopwords are manually removed, uninformative words common in that corpus will still dominate the most probable words in a topic. In this work, we first show how the standard topic quality measures of coherence and pointwise mutual information act counter-intuitively in the presence of common but irrelevant words, making it difficult to even quantitatively identify situations in which topics may be dominated by stopwords. We propose an additional topic quality metric that targets the stopword problem, and show that it, unlike the standard measures, correctly correlates with human judgements of quality. We also propose a simple-to-implement strategy for generating topics that are evaluated to be of much higher quality by both human assessment and our new metric. This approach, a collection of informative priors easily introduced into most LDA-style inference methods, automatically promotes terms with domain relevance and demotes domain-specific stop words. We demonstrate this approach's effectiveness in three very different domains: Department of Labor accident reports, online health forum posts, and NIPS abstracts. Overall we find that current practices thought to solve this problem do not do so adequately, and that our proposal offers a substantial improvement for those interested in interpreting their topics as objects in their own right.