Vulnerability to adversarial attacks is a well-known weakness of Deep Neural networks. While most of the studies focus on single-task neural networks with computer vision datasets, very little research has considered complex multi-task models that are common in real applications. In this paper, we evaluate the design choices that impact the robustness of multi-task deep learning networks. We provide evidence that blindly adding auxiliary tasks, or weighing the tasks provides a false sense of robustness. Thereby, we tone down the claim made by previous research and study the different factors which may affect robustness. In particular, we show that the choice of the task to incorporate in the loss function are important factors that can be leveraged to yield more robust models.

We extend the adversarial/non-stochastic multi-play multi-armed bandit (MPMAB) to the case where the number of arms to play is variable. The work is motivated by the fact that the resources allocated to scan different critical locations in an interconnected transportation system change dynamically over time and depending on the environment. By modeling the malicious hacker and the intrusion monitoring system as the attacker and the defender, respectively, we formulate the problem for the two players as a sequential pursuit-evasion game. We derive the condition under which a Nash equilibrium of the strategic game exists. For the defender side, we provide an exponential-weighted based algorithm with sublinear pseudo-regret. We further extend our model to heterogeneous rewards for both players, and obtain lower and upper bounds on the average reward for the attacker. We provide numerical experiments to demonstrate the effectiveness of a variable-arm play.

Goal-directed generation, aiming for solving downstream tasks by generating diverse data, has a potentially wide range of applications in the real world. Previous works tend to formulate goal-directed generation as a purely data-driven problem, which directly searches or approximates the distribution of samples satisfying the goal. However, the generation ability of preexisting work is heavily restricted by inefficient sampling, especially for sparse goals that rarely show up in off-the-shelf datasets. For instance, generating safety-critical traffic scenes with the goal of increasing the risk of collision is critical to evaluate autonomous vehicles, but the rareness of such scenes is the biggest resistance. In this paper, we integrate causality as a prior into the safety-critical scene generation process and propose a flow-based generative framework - Causal Autoregressive Flow (CausalAF). CausalAF encourages the generative model to uncover and follow the causal relationship among generated objects via novel causal masking operations instead of searching the sample only from observational data. By learning the cause-and-effect mechanism of how the generated scene achieves the goal rather than just learning correlations from data, CausalAF significantly improves the learning efficiency. Extensive experiments on three heterogeneous traffic scenes illustrate that CausalAF requires much fewer optimization resources to effectively generate goal-directed scenes for safety evaluation tasks.

Graph Neural Networks (GNNs) are increasingly important given their popularity and the diversity of applications. Yet, existing studies of their vulnerability to adversarial attacks rely on relatively small graphs. We address this gap and study how to attack and defend GNNs at scale. We propose two sparsity-aware first-order optimization attacks that maintain an efficient representation despite optimizing over a number of parameters which is quadratic in the number of nodes. We show that common surrogate losses are not well-suited for global attacks on GNNs. Our alternatives can double the attack strength. Moreover, to improve GNNs' reliability we design a robust aggregation function, Soft Median, resulting in an effective defense at all scales. We evaluate our attacks and defense with standard GNNs on graphs more than 100 times larger compared to previous work.

This paper studies the problem of performing a sequence of optimal interventions in a causal dynamical system where both the target variable of interest and the inputs evolve over time. This problem arises in a variety of domains e.g. system biology and operational research. Dynamic Causal Bayesian Optimization (DCBO) brings together ideas from sequential decision making, causal inference and Gaussian process (GP) emulation. DCBO is useful in scenarios where all causal effects in a graph are changing over time. At every time step DCBO identifies a local optimal intervention by integrating both observational and past interventional data collected from the system. We give theoretical results detailing how one can transfer interventional information across time steps and define a dynamic causal GP model which can be used to quantify uncertainty and find optimal interventions in practice. We demonstrate how DCBO identifies optimal interventions faster than competing approaches in multiple settings and applications.

Post-processing in algorithmic fairness is a versatile approach for correcting bias in ML systems that are already used in production. The main appeal of post-processing is that it avoids expensive retraining. In this work, we propose general post-processing algorithms for individual fairness (IF). We consider a setting where the learner only has access to the predictions of the original model and a similarity graph between individuals, guiding the desired fairness constraints. We cast the IF post-processing problem as a graph smoothing problem corresponding to graph Laplacian regularization that preserves the desired "treat similar individuals similarly" interpretation. Our theoretical results demonstrate the connection of the new objective function to a local relaxation of the original individual fairness. Empirically, our post-processing algorithms correct individual biases in large-scale NLP models such as BERT, while preserving accuracy.

In the hospital setting, a small percentage of recurrent frequent patients contribute to a disproportional amount of healthcare resource usage. Moreover, in many of these cases, patient outcomes can be greatly improved by reducing reoccurring visits, especially when they are associated with substance abuse, mental health, and medical factors that could be improved by social-behavioral interventions, outpatient or preventative care. To address this, we developed a computationally efficient and interpretable framework that both identifies recurrent patients with high utilization and determines which comorbidities contribute most to their recurrent visits. Specifically, we present a novel algorithm, called the minimum similarity association rules (MSAR), balancing confidence-support trade-off, to determine the conditions most associated with reoccurring Emergency department (ED) and inpatient visits. We validate MSAR on a large Electric Health Record (EHR) dataset. Part of the solution is deployed in Philips product Patient Flow Capacity Suite (PFCS).

U.S. officials say they believe Iran was behind the drone attack last week at the military outpost in southern Syria where American troops are based. Officials said Monday the U.S. believes that Iran resourced and encouraged the attack, but that the drones were not launched from Iran. They were Iranian drones, and Iran appears to have facilitated their use, officials said, speaking on condition of anonymity to discuss details that have not been made public. Officials said they believe the attacks involved as many as five drones laden with explosive charges, and that they hit both the U.S. side of al-Tanf garrison and the side where Syrian opposition forces stay. There were no reported injuries or deaths as a result of the attack.

United States officials believe Iran was behind a drone attack last week at a military outpost at al-Tanf in southern Syria where American troops are based. Officials said Monday the US believes that Iran provided resources and encouraged the attack, but that the drones were not launched from Iran, The Associated Press news service reported. The drones were Iranian, and Iran appears to have facilitated their use, officials said, speaking on condition of anonymity to discuss details that have not been made public. Officials said they believe the attacks involved as many as five drones laden with explosive charges, and that they hit both the US side of al-Tanf garrison and the side where Syrian opposition forces reside. There were no reported injuries or deaths as a result of the attack but it comes in a period of rising tensions between the US and Iran.

This post is to share with you the recent publication of the book: "Data Science for Economics and Finance: Methodologies and Applications", by Sergio Consoli, Diego Reforgiato Recupero, and Michaela Saisana. The use of data science and artificial intelligence for economics and finance is providing benefits for scientists, professionals and policy-makers by improving the available data analysis methodologies for economic forecasting and therefore making our societies better prepared for the challenges of tomorrow. This book is a good example of how combining expertise from the European Commission, universities in the U.S. and Europe, financial and economic institutions, and multilateral organizations, can bring forward a shared vision on the benefits of data science applied to economics and finance; from the research point of view to the evaluation of policies on the other hand. It showcases how data science is reshaping the business sector. It includes examples of novel big data sources and some successful applications on the use of advanced machine learning, natural language processing, networks analysis, and time series analysis and forecasting, among others, in the economic and financial sectors.