Learning invariant representations has been proposed as a key technique for addressing the domain generalization problem. However, the question of identifying the right conditions for invariance remains unanswered. In this work, we propose a causal interpretation of domain generalization that defines domains as interventions under a data-generating process. Based on a general causal model for data from multiple domains, we show that prior methods for learning an invariant representation optimize for an incorrect objective. We highlight an alternative condition: inputs across domains should have the same representation if they are derived from the same base object. In practice, knowledge about generation of data or objects is not available. Hence we propose an iterative algorithm called MatchDG that approximates base object similarity by using a contrastive loss formulation adapted for multiple domains. We then match inputs that are similar under the resultant representation to build an invariant classifier. We evaluate MatchDG on rotated MNIST, Fashion-MNIST, and PACS datasets and find that it outperforms prior work on out-of-domain accuracy and learns matches that have over 25\% overlap with ground-truth object matches in MNIST and Fashion-MNIST. Code repository can be accessed here: \textit{https://github.com/microsoft/robustdg}

Machine learning models, especially deep neural networks have been shown to reveal membership information of inputs in the training data. Such membership inference attacks are a serious privacy concern, for example, patients providing medical records to build a model that detects HIV would not want their identity to be leaked. Further, we show that the attack accuracy amplifies when the model is used to predict samples that come from a different distribution than the training set, which is often the case in real world applications. Therefore, we propose the use of causal learning approaches where a model learns the causal relationship between the input features and the outcome. Causal models are known to be invariant to the training distribution and hence generalize well to shifts between samples from the same distribution and across different distributions. First, we prove that models learned using causal structure provide stronger differential privacy guarantees than associational models under reasonable assumptions. Next, we show that causal models trained on sufficiently large samples are robust to membership inference attacks across different distributions of datasets and those trained on smaller sample sizes always have lower attack accuracy than corresponding associational models. Finally, we confirm our theoretical claims with experimental evaluation on $4$ datasets with moderately complex Bayesian networks. We observe that neural network-based associational models exhibit up to 80% attack accuracy under different test distributions and sample sizes whereas causal models exhibit attack accuracy close to a random guess. Our results confirm the value of the generalizability of causal models in reducing susceptibility to privacy attacks.

In critical decision-making scenarios, optimizing accuracy can lead to a biased classifier, hence past work recommends enforcing group-based fairness metrics in addition to maximizing accuracy. However, doing so exposes the classifier to another kind of bias called infra-marginality. This refers to individual-level bias where some individuals/subgroups can be worse off than under simply optimizing for accuracy. For instance, a classifier implementing race-based parity may significantly disadvantage women of the advantaged race. To quantify this bias, we propose a general notion of $\eta$-infra-marginality that can be used to evaluate the extent of this bias. We prove theoretically that, unlike other fairness metrics, infra-marginality does not have a trade-off with accuracy: high accuracy directly leads to low infra-marginality. This observation is confirmed through empirical analysis on multiple simulated and real-world datasets. Further, we find that maximizing group fairness often increases infra-marginality, suggesting the consideration of both group-level fairness and individual-level infra-marginality. However, measuring infra-marginality requires knowledge of the true distribution of individual-level outcomes correctly and explicitly. We propose a practical method to measure infra-marginality, and a simple algorithm to maximize group-wise accuracy and avoid infra-marginality.

Estimating average causal effect (ACE) is useful whenever we want to know the effect of an intervention on a given outcome. In the absence of a randomized experiment, many methods such as stratification and inverse propensity weighting have been proposed to estimate ACE. However, it is hard to know which method is optimal for a given dataset or which hyperparameters to use for a chosen method. To this end, we provide a framework to characterize the loss of a causal inference method against the true ACE, by framing causal inference as a representation learning problem. We show that many popular methods, including back-door methods can be considered as weighting or representation learning algorithms, and provide general error bounds for their causal estimates. In addition, we consider the case when unobserved variables can confound the causal estimate and extend proposed bounds using principles of robust statistics, considering confounding as contamination under the Huber contamination model. These bounds are also estimable; as an example, we provide empirical bounds for the Inverse Propensity Weighting (IPW) estimator and show how the bounds can be used to optimize the threshold of clipping extreme propensity scores. Our work provides a new way to reason about competing estimators, and opens up the potential of deriving new methods by minimizing the proposed error bounds.

Post-hoc explanations of machine learning models are crucial for people to understand and act on algorithmic predictions. An intriguing class of explanations is through counterfactuals, hypothetical examples that show people how to obtain a different prediction. We posit that effective counterfactual explanations should satisfy two properties: feasibility of the counterfactual actions given user context and constraints, and diversity among the counterfactuals presented. To this end, we propose a framework for generating and evaluating a diverse set of counterfactual explanations based on average distance and determinantal point processes. To evaluate the actionability of counterfactuals, we provide metrics that enable comparison of counterfactual-based methods to other local explanation methods. We further address necessary tradeoffs and point to causal implications in optimizing for counterfactuals. Our experiments on three real-world datasets show that our framework can generate a set of counterfactuals that are diverse and well approximate local decision boundaries.

Digital Adherence Technologies (DATs) are an increasingly popular method for verifying patient adherence to many medications. We analyze data from one city served by 99DOTS, a phone-call-based DAT deployed for Tuberculosis (TB) treatment in India where nearly 3 million people are afflicted with the disease each year. The data contains nearly 17,000 patients and 2.1M phone calls. We lay the groundwork for learning from this real-world data, including a method for avoiding the effects of unobserved interventions in training data used for machine learning. We then construct a deep learning model, demonstrate its interpretability, and show how it can be adapted and trained in three different clinical scenarios to better target and improve patient care. In the real-time risk prediction setting our model could be used to proactively intervene with 21% more patients and before 76% more missed doses than current heuristic baselines. For outcome prediction, our model performs 40% better than baseline methods, allowing cities to target more resources to clinics with a heavier burden of patients at risk of failure. Finally, we present a case study demonstrating how our model can be trained in an end-to-end decision focused learning setting to achieve 15% better solution quality in an example decision problem faced by health workers.

The Workshop Program of the Association for the Advancement of Artificial Intelligence’s 12th International Conference on Web and Social Media (AAAI-18) was held at Stanford University, Stanford, California USA, on Monday, June 25, 2018. There were fourteen workshops in the program: Algorithmic Personalization and News: Risks and Opportunities; Beyond Online Data: Tackling Challenging Social Science Questions; Bridging the Gaps: Social Media, Use and Well-Being; Chatbot; Data-Driven Personas and Human-Driven Analytics: Automating Customer Insights in the Era of Social Media;  Designed Data for Bridging the Lab and the Field: Tools, Methods, and Challenges in Social Media Experiments; Emoji Understanding and Applications in Social Media; Event Analytics Using Social Media Data; Exploring Ethical Trade-Offs in Social Media Research; Making Sense of Online Data for Population Research; News and Public Opinion; Social Media and Health: A Focus on Methods for Linking Online and Offline Data; Social Web for Environmental and Ecological Monitoring and The ICWSM Science Slam. Workshops were held on the first day of the conference. Workshop participants met and discussed issues with a selected focus — providing an informal setting for active exchange among researchers, developers, and users on topics of current interest. Organizers from nine of the  workshops submitted reports, which are reproduced in this report. Brief summaries of the other five workshops have been reproduced from their website descriptions.

We present a method for estimating causal effects in time series data when fine-grained information about the outcome of interest is available. Specifically, we examine what we call the split-door setting, where the outcome variable can be split into two parts: one that is potentially affected by the cause being studied and another that is independent of it, with both parts sharing the same (unobserved) confounders. We show that under these conditions, the problem of identification reduces to that of testing for independence among observed variables, and present a method that uses this approach to automatically find subsets of the data that are causally identified. We demonstrate the method by estimating the causal impact of Amazon's recommender system on traffic to product pages, finding thousands of examples within the dataset that satisfy the split-door criterion. Unlike past studies based on natural experiments that were limited to a single product category, our method applies to a large and representative sample of products viewed on the site. In line with previous work, we find that the widely-used click-through rate (CTR) metric overestimates the causal impact of recommender systems; depending on the product category, we estimate that 50-80\% of the traffic attributed to recommender systems would have happened even without any recommendations. We conclude with guidelines for using the split-door criterion as well as a discussion of other contexts where the method can be applied.

Two main approaches to using social network information in recommendation have emerged: augmenting collaborative filtering with social data and algorithms that use only ego-centric data. We compare the two approaches using movie and music data from Facebook, and hashtag data from Twitter. We find that recommendation algorithms based only on friends perform no worse than those based on the full network, even though they require much less data and computational resources. Further, our evidence suggests that locality of preference, or the non-random distribution of item preferences in a social network, is a driving force behind the value of incorporating social network information into recommender algorithms. When locality is high, as in Twitter data, simple k-nn recommenders do better based only on friends than they do if they draw from the entire network. These results help us understand when, and why, social network information is likely to support recommendation systems, and show that systems that see ego-centric slices of a complete network (such as websites that use Facebook logins) or have computational limitations (such as mobile devices) may profitably use ego-centric recommendation algorithms.