The importance of studying the robustness of learners to malicious data is well established. While much work has been done establishing both robust estimators and effective data injection attacks when the attacker is omniscient, the ability of an attacker to provably harm learning while having access to little information is largely unstudied. We study the potential of a "blind attacker" to provably limit a learner's performance by data injection attack without observing the learner's training set or any parameter of the distribution from which it is drawn. We provide examples of simple yet effective attacks in two settings: firstly, where an "informed learner" knows the strategy chosen by the attacker, and secondly, where a "blind learner" knows only the proportion of malicious data and some family to which the malicious distribution chosen by the attacker belongs. For each attack, we analyze minimax rates of convergence and establish lower bounds on the learner's minimax risk, exhibiting limits on a learner's ability to learn under data injection attack even when the attacker is "blind".

LPAR 2015: 372-386 [7] Robert Veroff: Using Hints to Increase the Effectiveness of an Automated Reasoning Program: Case Studies.

The foundation model enables general-purpose problem-solving and enjoys desirable rapid adaptation due to its adopted cross-task generalization paradigms, e.g., pretraining, meta-training, and finetuning. Recent advances in these paradigms show the crucial role of challenging tasks' prioritized sampling in enhancing adaptation robustness. However, ranking task difficulties exhausts massive task queries to evaluate, thus computation and annotation intensive, which is typically unaffordable in practice. This work underscores the criticality of both adaptation robustness and learning efficiency, especially in scenarios where tasks are risky or costly to evaluate, e.g., policy evaluations in Markov decision processes (MDPs) or inference with large models. To this end, we present Model Predictive Task Sampling (MPTS) to establish connections between the task space and adaptation risk landscape to form a theoretical guideline in robust active task sampling. MPTS characterizes the task episodic information with a generative model and directly predicts task-specific adaptation risk values from posterior inference. The developed risk learner can amortize expensive evaluation and provably approximately rank task difficulties in the pursuit of task robust adaptation. MPTS can be seamlessly integrated into zero-shot, few-shot, and many-shot learning paradigms. Extensive experimental results are conducted to exhibit the superiority of the proposed framework, remarkably increasing task adaptation robustness and retaining learning efficiency in contrast to existing state-of-the-art (SOTA) methods. The code is available at the project site https://github.com/thu-rllab/MPTS.

Overall, I find this branch of work quite interesting and am glad the authors are choosing to study this problem. The attacks mentioned in the paper may become feasible in the age of large web-scale datasets or human-in-the-loop training systems, along with the privacy scenario mentioned in the paper. The authors do an excellent job of motivating the problem. The paper appears clearly written if the reader is an expert within the field of statistical decision theory. I must admit that this is not my area of expertise.

Machine learning techniques are widely used for estimating causal effects. Double/debiased machine learning (DML) (Chernozhukov et al., 2018) uses a double-robust score function that relies on the prediction of nuisance functions, such as the propensity score, which is the probability of treatment assignment conditional on covariates. Estimators relying on double-robust score functions are highly sensitive to errors in propensity score predictions. Machine learners increase the severity of this problem as they tend to over- or underestimate these probabilities. Several calibration approaches have been proposed to improve probabilistic forecasts of machine learners. This paper investigates the use of probability calibration approaches within the DML framework. Simulation results demonstrate that calibrating propensity scores may significantly reduces the root mean squared error of DML estimates of the average treatment effect in finite samples. We showcase it in an empirical example and provide conditions under which calibration does not alter the asymptotic properties of the DML estimator.

We introduce the package ddml for Double/Debiased Machine Learning (DDML) in Stata. Estimators of causal parameters for five different econometric models are supported, allowing for flexible estimation of causal effects of endogenous variables in settings with unknown functional forms and/or many exogenous variables. ddml is compatible with many existing supervised machine learning programs in Stata. We recommend using DDML in combination with stacking estimation which combines multiple machine learners into a final predictor. We provide Monte Carlo evidence to support our recommendation.

The causal inference literature frequently focuses on estimating the mean of the potential outcome, whereas the quantiles of the potential outcome may carry important additional information. We propose a universal approach, based on the inverse estimating equations, to generalize a wide class of causal inference solutions from estimating the mean of the potential outcome to its quantiles. We assume that an identifying moment function is available to identify the mean of the threshold-transformed potential outcome, based on which a convenient construction of the estimating equation of quantiles of potential outcome is proposed. In addition, we also give a general construction of the efficient influence functions of the mean and quantiles of potential outcomes, and identify their connection. We motivate estimators for the quantile estimands with the efficient influence function, and develop their asymptotic properties when either parametric models or data-adaptive machine learners are used to estimate the nuisance functions. A broad implication of our results is that one can rework the existing result for mean causal estimands to facilitate causal inference on quantiles, rather than starting from scratch. Our results are illustrated by several examples.

pystacked implements stacked generalization (Wolpert, 1992) for regression and binary classification via Python's scikit-learn. Stacking combines multiple supervised machine learners -- the "base" or "level-0" learners -- into a single learner. The currently supported base learners include regularized regression, random forest, gradient boosted trees, support vector machines, and feed-forward neural nets (multi-layer perceptron). pystacked can also be used with as a `regular' machine learning program to fit a single base learner and, thus, provides an easy-to-use API for scikit-learn's machine learning algorithms.

The Jobs and Skills Summit being convened this week in Canberra sets out to address the employment challenges that Australia currently faces. But even as we respond to today's skills shortage, we must also keep an eye on the long-term consequences of Artificial Intelligence (AI) and machine learning on jobs. Fantasies in which AIs advance to the point of doing all of our jobs are seductive in some tech circles. Elon Musk has long promised fully driverless cars that will be safer than cars driven by humans. When these vehicles do arrive, they may delete the human Uber driver.

Learning how to learn is something most humans do well, by leveraging previous experiences to inform the learning processes for new tasks. Endowing AI systems with such abilities however remains challenging, as it requires the machine learners to learn update rules, which typically have been manually tuned for each task. The field of meta-learning studies how to enable machine learners to learn how to learn, and is a critical research area for improving the efficiency of AI agents. One of the approaches is for learners to learn an update rule by applying it on previous steps and then evaluating the corresponding performance. To fully unlock the potential of meta-learning, it is necessary to overcome both the meta-optimization problem and myopic meta objectives.