Search for: All records

The use of AI-based decision aids in diverse domains has inspired many empirical investigations into how AI models’ decision recommendations impact humans’ decision accuracy in AI-assisted decision making, while explorations on the impacts on humans’ decision fairness are largely lacking despite their clear importance. In this paper, using a real-world business decision making scenario—bidding in rental housing markets—as our testbed, we present an experimental study on understanding how the bias level of the AI-based decision aid as well as the provision of AI explanations affect the fairness level of humans’ decisions, both during and after their usage of the decision aid. Our results suggest that when people are assisted by an AI-based decision aid, both the higher level of racial biases the decision aid exhibits and surprisingly, the presence of AI explanations, result in more unfair human decisions across racial groups. Moreover, these impacts are partly made through triggering humans’ “disparate interactions” with AI. However, regardless of the AI bias level and the presence of AI explanations, when people return to make independent decisions after their usage of the AI-based decision aid, their decisions no longer exhibit significant unfairness across racial groups.

 

We study linear bandits when the underlying reward function is not linear. Existing work relies on a uniform misspecification parameter $\epsilon$ that measures the sup-norm error of the best linear approximation. This results in an unavoidable linear regret whenever $\epsilon > 0$. We describe a more natural model of misspecification which only requires the approximation error at each input $x$ to be proportional to the suboptimality gap at $x$. It captures the intuition that, for optimization problems, near-optimal regions should matter more and we can tolerate larger approximation errors in suboptimal regions. Quite surprisingly, we show that the classical LinUCB algorithm — designed for the realizable case — is automatically robust against such gap-adjusted misspecification. It achieves a near-optimal $\sqrt{T}$ regret for problems that the best-known regret is almost linear in time horizon $T$. Technically, our proof relies on a novel self-bounding argument that bounds the part of the regret due to misspecification by the regret itself. 

The increased integration of artificial intelligence (AI) technologies in human workflows has resulted in a new paradigm of AI-assisted decision making,in which an AI model provides decision recommendations while humans make the final decisions. To best support humans in decision making, it is critical to obtain a quantitative understanding of how humans interact with and rely on AI. Previous studies often model humans' reliance on AI as an analytical process, i.e., reliance decisions are made based on cost-benefit analysis. However, theoretical models in psychology suggest that the reliance decisions can often be driven by emotions like humans' trust in AI models. In this paper, we propose a hidden Markov model to capture the affective process underlying the human-AI interaction in AI-assisted decision making, by characterizing how decision makers adjust their trust in AI over time and make reliance decisions based on their trust. Evaluations on real human behavior data collected from human-subject experiments show that the proposed model outperforms various baselines in accurately predicting humans' reliance behavior in AI-assisted decision making. Based on the proposed model, we further provide insights into how humans' trust and reliance dynamics in AI-assisted decision making is influenced by contextual factors like decision stakes and their interaction experiences. 

General function approximation is a powerful tool to handle large state and action spaces in a broad range of reinforcement learning (RL) scenarios. However, theoretical understanding of non-stationary MDPs with general function approximation is still limited. In this paper, we make the first such an attempt. We first propose a new complexity metric called dynamic Bellman Eluder (DBE) dimension for non-stationary MDPs, which subsumes majority of existing tractable RL problems in static MDPs as well as non-stationary MDPs. Based on the proposed complexity metric, we propose a novel confidence-set based model-free algorithm called SW-OPEA, which features a sliding window mechanism and a new confidence set design for non-stationary MDPs. We then establish an upper bound on the dynamic regret for the proposed algorithm, and show that SW-OPEA is provably efficient as long as the variation budget is not significantly large. We further demonstrate via examples of non-stationary linear and tabular MDPs that our algorithm performs better in small variation budget scenario than the existing UCB-type algorithms. To the best of our knowledge, this is the first dynamic regret analysis in non-stationary MDPs with general function approximation. 

Behavior constrained policy optimization has been demonstrated to be a successful paradigm for tackling Offline Reinforcement Learning. By exploiting historical transitions, a policy is trained to maximize a learned value function while constrained by the behavior policy to avoid a significant distributional shift. In this paper, we propose our closed-form policy improvement operators. We make a novel observation that the behavior constraint naturally motivates the use of first-order Taylor approximation, leading to a linear approximation of the policy objective. Additionally, as practical datasets are usually collected by heterogeneous policies, we model the behavior policies as a Gaussian Mixture and overcome the induced optimization difficulties by leveraging the LogSumExp’s lower bound and Jensen’s Inequality, giving rise to a closed-form policy improvement operator. We instantiate both one-step and iterative offline RL algorithms with our novel policy improvement operators and empirically demonstrate their effectiveness over state-of-the-art algorithms on the standard D4RL benchmark. Our code is available at https://cfpi-icml23.github.io/. 

Offline reinforcement learning, which aims at optimizing sequential decision-making strategies with historical data, has been extensively applied in real-life applications. State-Of-The-Art algorithms usually leverage powerful function approximators (e.g. neural networks) to alleviate the sample complexity hurdle for better empirical performances. Despite the successes, a more systematic under- standing of the statistical complexity for function approximation remains lacking. Towards bridging the gap, we take a step by considering offline reinforcement learning with differentiable function class approximation (DFA). This function class naturally incorporates a wide range of models with nonlinear/nonconvex structures. We show offline RL with differentiable function approximation is provably efficient by analyzing the pessimistic fitted Q-learning (PFQL) algorithm, and our results provide the theoretical basis for understanding a variety of practical heuristics that rely on Fitted Q-Iteration style design. In addition, we further im- prove our guarantee with a tighter instance-dependent characterization. We hope our work could draw interest in studying reinforcement learning with differentiable function approximation beyond the scope of current research. 

Recent years have witnessed the growing literature in empirical evaluation of explainable AI (XAI) methods. This study contributes to this ongoing conversation by presenting a comparison on the effects of a set of established XAI methods in AI-assisted decision making. Based on our review of previous literature, we highlight three desirable properties that ideal AI explanations should satisfy — improve people’s understanding of the AI model, help people recognize the model uncertainty, and support people’s calibrated trust in the model. Through three randomized controlled experiments, we evaluate whether four types of common model-agnostic explainable AI methods satisfy these properties on two types of AI models of varying levels of complexity, and in two kinds of decision making contexts where people perceive themselves as having different levels of domain expertise. Our results demonstrate that many AI explanations do not satisfy any of the desirable properties when used on decision making tasks that people have little domain expertise in. On decision making tasks that people are more knowledgeable, the feature contribution explanation is shown to satisfy more desiderata of AI explanations, even when the AI model is inherently complex. We conclude by discussing the implications of our study for improving the design of XAI methods to better support human decision making, and for advancing more rigorous empirical evaluation of XAI methods.