The prevalence of machine learning models in various industries has led to growing demands for model interpretability and for the ability to provide meaningful recourse to users. For example, patients hoping to improve their diagnoses or loan applicants seeking to increase their chances of approval. Counterfactuals can help in this regard by identifying input perturbations that would result in more desirable prediction outcomes. Meaningful counterfactuals should be able to achieve the desired outcome, but also be realistic, actionable, and efficient to compute. Current approaches achieve desired outcomes with moderate actionability but are severely limited in terms of realism and latency. To tackle these limitations, we apply Generative Adversarial Nets (GANs) toward counterfactual search. We also introduce a novel Residual GAN (RGAN) that helps to improve counterfactual realism and actionability compared to regular GANs. The proposed CounteRGAN method utilizes an RGAN and a target classifier to produce counterfactuals capable of providing meaningful recourse. Evaluations on two popular datasets highlight how the CounteRGAN is able to overcome the limitations of existing methods, including latency improvements of >50x to >90,000x, making meaningful recourse available in real-time and applicable to a wide range of domains.
Many recent studies in deep reinforcement learning (DRL) have proposed to boost adversarial robustness through policy distillation utilizing adversarial training, where additional adversarial examples are added in the training process of the student policy; this makes the robustness improvement less flexible and more computationally expensive. In contrast, we propose an efficient policy distillation paradigm called robust policy distillation that is capable of achieving an adversarially robust student policy without relying on any adversarial example during student policy training. To this end, we devise a new policy distillation loss that consists of two terms: 1) a prescription gap maximization loss aiming at simultaneously maximizing the likelihood of the action selected by the teacher policy and the entropy over the remaining actions; 2) a Jacobian regularization loss that minimizes the magnitude of Jacobian with respect to the input state. The theoretical analysis proves that our distillation loss guarantees to increase the prescription gap and the adversarial robustness. Meanwhile, experiments on five Atari games firmly verifies the superiority of our policy distillation on boosting adversarial robustness compared to other state-of-the-arts.
Contact tracing has grown in popularity as a promising solution to the COVID-19 pandemic. The benefits of automated contact tracing are two-fold. Contact tracing promises to reduce the number of infections by being able to: 1) systematically identify all of those that have been in contact with someone who has had COVID; and, 2) ensure those that have been exposed to the virus do not unknowingly infect others. "COVI" is the name of a recent contact tracing app developed by Mila and was proposed to help combat COVID-19 in Canada. The app was designed to inform each individual of their relative risk of being infected with the virus, which Mila claimed would empower citizens to make informed decisions about their movement and allow for a data-driven approach to public health policy; all the while ensuring data is safeguarded from governments, companies, and individuals. This article will provide a critical response to Mila's COVI White Paper. Specifically, this article will discuss: the extent to which diversity has been considered in the design of the app, assumptions surrounding users' interaction with the app and the app's utility, as well as unanswered questions surrounding transparency, accountability, and security. We see this as an opportunity to supplement the excellent risk analysis done by the COVI team to surface insights that can be applied to other contact- and proximity-tracing apps that are being developed and deployed across the world. Our hope is that, through a meaningful dialogue, we can ultimately help organizations develop better solutions that respect the fundamental rights and values of the communities these solutions are meant to serve.
Security and ethics are both core to ensuring that a machine learning system can be trusted. In production machine learning, there is generally a hand-off from those who build a model to those who deploy a model. In this hand-off, the engineers responsible for model deployment are often not privy to the details of the model and thus, the potential vulnerabilities associated with its usage, exposure, or compromise. Techniques such as model theft, model inversion, or model misuse may not be considered in model deployment, and so it is incumbent upon data scientists and machine learning engineers to understand these potential risks so they can communicate them to the engineers deploying and hosting their models. This is an open problem in the machine learning community and in order to help alleviate this issue, automated systems for validating privacy and security of models need to be developed, which will help to lower the burden of implementing these hand-offs and increasing the ubiquity of their adoption.
A fundamental assumption of most machine learning algorithms is that the training and test data are drawn from the same underlying distribution. However, this assumption is violated in almost all practical applications: machine learning systems are regularly tested on data that are structurally different from the training set, either due to temporal correlations, particular end users, or other factors. In this work, we consider the setting where test examples are not drawn from the training distribution. Prior work has approached this problem by attempting to be robust to all possible test time distributions, which may degrade average performance, or by "peeking" at the test examples during training, which is not always feasible. In contrast, we propose to learn models that are adaptable, such that they can adapt to distribution shift at test time using a batch of unlabeled test data points. We acquire such models by learning to adapt to training batches sampled according to different sub-distributions, which simulate structural distribution shifts that may occur at test time. We introduce the problem of adaptive risk minimization (ARM), a formalization of this setting that lends itself to meta-learning methods. Compared to a variety of methods under the paradigms of empirical risk minimization and robust optimization, our approach provides substantial empirical gains on image classification problems in the presence of distribution shift.
These past few months have been especially challenging, and the deployment of technology in ways hitherto untested at an unrivalled pace has left the internet and technology watchers aghast. Artificial intelligence has become the byword for technological progress and is being used in everything from helping us combat the COVID-19 pandemic to nudging our attention in different directions as we all spend increasingly larger amounts of time online. It has never been more important that we keep a sharp eye out on the development of this field and how it is shaping our society and interactions with each other. With this inaugural edition of the State of AI Ethics we hope to bring forward the most important developments that caught our attention at the Montreal AI Ethics Institute this past quarter. Our goal is to help you navigate this ever-evolving field swiftly and allow you and your organization to make informed decisions. This pulse-check for the state of discourse, research, and development is geared towards researchers and practitioners alike who are making decisions on behalf of their organizations in considering the societal impacts of AI-enabled solutions. We cover a wide set of areas in this report spanning Agency and Responsibility, Security and Risk, Disinformation, Jobs and Labor, the Future of AI Ethics, and more. Our staff has worked tirelessly over the past quarter surfacing signal from the noise so that you are equipped with the right tools and knowledge to confidently tread this complex yet consequential domain.
We propose two novel stochastic gradient descent algorithms, ByGARS and ByGARS++, for distributed machine learning in the presence of Byzantine adversaries. In these algorithms, reputation score of workers are computed using an auxiliary dataset with a larger stepsize. This reputation score is then used for aggregating the gradients for stochastic gradient descent with a smaller stepsize. We show that using these reputation scores for gradient aggregation is robust to any number of Byzantine adversaries. In contrast to prior works targeting any number of adversaries, we improve the generalization performance by making use of some adversarial workers along with the benign ones. The computational complexity of ByGARS++ is the same as the usual stochastic gradient descent method with only an additional inner product computation. We establish its convergence for strongly convex loss functions and demonstrate the effectiveness of the algorithms for non-convex learning problems using MNIST and CIFAR-10 datasets.
Much of the current work on reinforcement learning studies episodic settings, where the agent is reset between trials to an initial state distribution, often with well-shaped reward functions. Non-episodic settings, where the agent must learn through continuous interaction with the world without resets, and where the agent receives only delayed and sparse reward signals, is substantially more difficult, but arguably more realistic considering real-world environments do not present the learner with a convenient "reset mechanism" and easy reward shaping. In this paper, instead of studying algorithmic improvements that can address such non-episodic and sparse reward settings, we instead study the kinds of environment properties that can make learning under such conditions easier. Understanding how properties of the environment impact the performance of reinforcement learning agents can help us to structure our tasks in ways that make learning tractable. We first discuss what we term "environment shaping" -- modifications to the environment that provide an alternative to reward shaping, and may be easier to implement. We then discuss an even simpler property that we refer to as "dynamism," which describes the degree to which the environment changes independent of the agent's actions and can be measured by environment transition entropy. Surprisingly, we find that even this property can substantially alleviate the challenges associated with non-episodic RL in sparse reward settings. We provide an empirical evaluation on a set of new tasks focused on non-episodic learning with sparse rewards. Through this study, we hope to shift the focus of the community towards analyzing how properties of the environment can affect learning and the ultimate type of behavior that is learned via RL.
In February 2020, the European Commission (EC) published a white paper entitled, On Artificial Intelligence - A European approach to excellence and trust. This paper outlines the EC's policy options for the promotion and adoption of artificial intelligence (AI) in the European Union. The Montreal AI Ethics Institute (MAIEI) reviewed this paper and published a response addressing the EC's plans to build an "ecosystem of excellence" and an "ecosystem of trust," as well as the safety and liability implications of AI, the internet of things (IoT), and robotics. MAIEI provides 15 recommendations in relation to the sections outlined above, including: 1) focus efforts on the research and innovation community, member states, and the private sector; 2) create alignment between trading partners' policies and EU policies; 3) analyze the gaps in the ecosystem between theoretical frameworks and approaches to building trustworthy AI; 4) focus on coordination and policy alignment; 5) focus on mechanisms that promote private and secure sharing of data; 6) create a network of AI research excellence centres to strengthen the research and innovation community; 7) promote knowledge transfer and develop AI expertise through Digital Innovation Hubs; 8) add nuance to the discussion regarding the opacity of AI systems; 9) create a process for individuals to appeal an AI system's decision or output; 10) implement new rules and strengthen existing regulations; 11) ban the use of facial recognition technology; 12) hold all AI systems to similar standards and compulsory requirements; 13) ensure biometric identification systems fulfill the purpose for which they are implemented; 14) implement a voluntary labelling system for systems that are not considered high-risk; 15) appoint individuals to the oversight process who understand AI systems well and are able to communicate potential risks.
Reinforcement learning provides an appealing formalism for learning control policies from experience. However, the classic active formulation of reinforcement learning necessitates a lengthy active exploration process for each behavior, making it difficult to apply in real-world settings. If we can instead allow reinforcement learning to effectively use previously collected data to aid the online learning process, where the data could be expert demonstrations or more generally any prior experience, we could make reinforcement learning a substantially more practical tool. While a number of recent methods have sought to learn offline from previously collected data, it remains exceptionally difficult to train a policy with offline data and improve it further with online reinforcement learning. In this paper we systematically analyze why this problem is so challenging, and propose a novel algorithm that combines sample-efficient dynamic programming with maximum likelihood policy updates, providing a simple and effective framework that is able to leverage large amounts of offline data and then quickly perform online fine-tuning of reinforcement learning policies. We show that our method enables rapid learning of skills with a combination of prior demonstration data and online experience across a suite of difficult dexterous manipulation and benchmark tasks.