The path toward realizing the potential of seasonal forecasting and its socioeconomic benefits depends heavily on improving general circulation model based dynamical forecasting systems. To improve dynamical seasonal forecast, it is crucial to set up forecast benchmarks, and clarify forecast limitations posed by model initialization errors, formulation deficiencies, and internal climate variability. With huge cost in generating large forecast ensembles, and limited observations for forecast verification, the seasonal forecast benchmarking and diagnosing task proves challenging. In this study, we develop a probabilistic deep neural network model, drawing on a wealth of existing climate simulations to enhance seasonal forecast capability and forecast diagnosis. By leveraging complex physical relationships encoded in climate simulations, our probabilistic forecast model demonstrates favorable deterministic and probabilistic skill compared to state-of-the-art dynamical forecast systems in quasi-global seasonal forecast of precipitation and near-surface temperature. We apply this probabilistic forecast methodology to quantify the impacts of initialization errors and model formulation deficiencies in a dynamical seasonal forecasting system. We introduce the saliency analysis approach to efficiently identify the key predictors that influence seasonal variability. Furthermore, by explicitly modeling uncertainty using variational Bayes, we give a more definitive answer to how the El Nino/Southern Oscillation, the dominant mode of seasonal variability, modulates global seasonal predictability.

We introduce the problem of regret minimization in Adversarial Dueling Bandits. As in classic Dueling Bandits, the learner has to repeatedly choose a pair of items and observe only a relative binary `win-loss' feedback for this pair, but here this feedback is generated from an arbitrary preference matrix, possibly chosen adversarially. Our main result is an algorithm whose $T$-round regret compared to the \emph{Borda-winner} from a set of $K$ items is $\tilde{O}(K^{1/3}T^{2/3})$, as well as a matching $\Omega(K^{1/3}T^{2/3})$ lower bound. We also prove a similar high probability regret bound. We further consider a simpler \emph{fixed-gap} adversarial setup, which bridges between two extreme preference feedback models for dueling bandits: stationary preferences and an arbitrary sequence of preferences. For the fixed-gap adversarial setup we give an $\smash{ \tilde{O}((K/\Delta^2)\log{T}) }$ regret algorithm, where $\Delta$ is the gap in Borda scores between the best item and all other items, and show a lower bound of $\Omega(K/\Delta^2)$ indicating that our dependence on the main problem parameters $K$ and $\Delta$ is tight (up to logarithmic factors).

Over the past years, bandit algorithms have found application in computational advertising, recommender systems, clinical trials, and many more. These algorithms make online decisions by balancing between exploiting previously high-reward actions vs. exploring less known ones that could potentially lead to higher rewards. Bandit problems can roughly be categorized [17] into stochastic bandits, in which subsequently played actions yield independent rewards, and adversarial bandits, where the rewards are chosen by an adversary, possibly subject to constraints. A recent line of works has sought to reap the benefits of both approaches by studying bandit problems that are stochastic in nature, but with rewards subject to a limited amount of adversarial corruption. Various works have developed provably robust algorithms [4, 11, 20, 23], and attacks have been designed that cause standard algorithms to fail [9, 11, 12, 21]. While near-optimal theoretical guarantees have been established in the case of independent arms [11], more general settings remain relatively poorly understood or even entirely unexplored; see Section 1.2 for details. Our primary goal is to bridge these gaps via a detailed study of stochastic linear bandits with adversarial corruptions. In the case of a fixed finite (but possibly very large) set of arms, we develop an elimination-based robust algorithm and provide regret bounds with a near-optimal joint dependence on the time horizon and the adversarial attack budget, demonstrating distinct behavior depending on whether the attack budget is known or unknown. In addition, we introduce a novel contextual linear bandit setting under adversarial corruptions, and show that under a context diversity assumption, a simple greedy algorithm attains near-optimal regret under adversarial corruptions, despite having no built-in mechanism that explicitly encourages exploration or robustness.

Autonomous spacecraft maneuver planning using an evolutionary algorithmic approach is investigated. Simulated spacecraft were placed into four different initial orbits. Each was allowed a string of thirty delta-v impulse maneuvers in six cartesian directions, the positive and negative x, y and z directions. The goal of the spacecraft maneuver string was to, starting from some non-polar starting orbit, place the spacecraft into a polar, low eccentricity orbit. A genetic algorithm was implemented, using a mating, fitness, mutation and crossover scheme for impulse strings. The genetic algorithm was successfully able to produce this result for all the starting orbits. Performance and future work is also discussed.

The challenge of establishing assurance in autonomy is rapidly attracting increasing interest in the industry, government, and academia. Autonomy is a broad and expansive capability that enables systems to behave without direct control by a human operator. To that end, it is expected to be present in a wide variety of systems and applications. A vast range of industrial sectors, including (but by no means limited to) defense, mobility, health care, manufacturing, and civilian infrastructure, are embracing the opportunities in autonomy yet face the similar barriers toward establishing the necessary level of assurance sooner or later. Numerous government agencies are poised to tackle the challenges in assured autonomy. Given the already immense interest and investment in autonomy, a series of workshops on Assured Autonomy was convened to facilitate dialogs and increase awareness among the stakeholders in the academia, industry, and government. This series of three workshops aimed to help create a unified understanding of the goals for assured autonomy, the research trends and needs, and a strategy that will facilitate sustained progress in autonomy. The first workshop, held in October 2019, focused on current and anticipated challenges and problems in assuring autonomous systems within and across applications and sectors. The second workshop held in February 2020, focused on existing capabilities, current research, and research trends that could address the challenges and problems identified in workshop. The third event was dedicated to a discussion of a draft of the major findings from the previous two workshops and the recommendations.

In Machine Learning (ML) models used for supporting decisions in high-stakes domains such as public policy, explainability is crucial for adoption and effectiveness. While the field of explainable ML has expanded in recent years, much of this work does not take real-world needs into account. A majority of proposed methods use benchmark ML problems with generic explainability goals without clear use-cases or intended end-users. As a result, the effectiveness of this large body of theoretical and methodological work on real-world applications is unclear. This paper focuses on filling this void for the domain of public policy. We develop a taxonomy of explainability use-cases within public policy problems; for each use-case, we define the end-users of explanations and the specific goals explainability has to fulfill; third, we map existing work to these use-cases, identify gaps, and propose research directions to fill those gaps in order to have practical policy impact through ML.

Continuous observation of climate indicators, such as trends in lake freezing, is important to understand the dynamics of the local and global climate system. Consequently, lake ice has been included among the Essential Climate Variables (ECVs) of the Global Climate Observing System (GCOS), and there is a need to set up operational monitoring capabilities. Multi-temporal satellite images and publicly available webcam streams are among the viable data sources to monitor lake ice. In this work we investigate machine learning-based image analysis as a tool to determine the spatio-temporal extent of ice on Swiss Alpine lakes as well as the ice-on and ice-off dates, from both multispectral optical satellite images (VIIRS and MODIS) and RGB webcam images. We model lake ice monitoring as a pixel-wise semantic segmentation problem, i.e., each pixel on the lake surface is classified to obtain a spatially explicit map of ice cover. We show experimentally that the proposed system produces consistently good results when tested on data from multiple winters and lakes. Our satellite-based method obtains mean Intersection-over-Union (mIoU) scores >93%, for both sensors. It also generalises well across lakes and winters with mIoU scores >78% and >80% respectively. On average, our webcam approach achieves mIoU values of 87% (approx.) and generalisation scores of 71% (approx.) and 69% (approx.) across different cameras and winters respectively. Additionally, we put forward a new benchmark dataset of webcam images (Photi-LakeIce) which includes data from two winters and three cameras.

The lack of interpretability and transparency are preventing economists from using advanced tools like neural networks in their empirical research. In this paper, we propose a class of interpretable neural network models that can achieve both high prediction accuracy and interpretability. The model can be written as a simple function of a regularized number of interpretable features, which are outcomes of interpretable functions encoded in the neural network. Researchers can design different forms of interpretable functions based on the nature of their tasks. In particular, we encode a class of interpretable functions named persistent change filters in the neural network to study time series cross-sectional data. We apply the model to predicting individual's monthly employment status using highdimensional administrative data. We achieve an accuracy of 94.5% in the test set, which is comparable to the best performed conventional machine learning methods. Furthermore, the interpretability of the model allows us to understand the mechanism that underlies the prediction: an individual's employment status is closely related to whether she pays different types of insurances. Our work is a useful step towards overcoming the "black box" problem of neural networks, and provide a new tool for economists to study administrative and proprietary big data.

Surveillance is becoming an increasingly controversial application given the rapid pace at which AI systems are being developed and deployed worldwide. While protestors marched through the city demanding justice for George Floyd and an end to police brutality, Minneapolis police trained surveillance tools to identify them. With just hours to sift through thousands of CCTV camera feeds and other dragnet data streams, the police turned to a range of automated systems for help, reaching for information collected by automated license plate readers, CCTV-video analysis software, open-source geolocation tools, and Clearview AI's controversial facial recognition system. High above the city, an unarmed Predator drone flew in circles, outfitted with a specialized camera first pioneered by the police in Baltimore that is capable of identifying individuals from 10,000 feet in the air, providing real-time surveillance of protestors across the city. But Minneapolis is not an isolated case of excessive policing and technology run amok. Instead, it is part of a larger strategy by the state, local, and federal government to build surveillance dragnets that pull in people's emails, texts, bank records, and smartphone location as well as their faces, movements, and physical whereabouts to equip law enforcement with unprecedented tools to search for and identify Americans without a warrant.

In this article, Juan Murillo, Senior Manager of Data Strategy at BBVA, and Jesús Lozano, Manager of Digital Regulation at BBVA, analyse the potential implications of Artificial Intelligence regulations and share their insights into the considerations that should be taken into account to ensure that regulatory aspects support the proper development of this discipline in the future. Artificial Intelligence is a term coined in the 1950s that is usually understood as referring to a single technology, when in reality it encompasses a broad range of techniques and methodologies whose theoretical foundations were laid over 70 years ago. This field has already gone through a number of stages. During the first stage, symbolic AI applications dominated. Symbolic AI is a top-down approach that aspires to parameterise all the alternatives to a problem in order to find the right solution by following a tree of logical rules.