We propose training fitted Q-iteration with log-loss (FQI-LOG) for batch reinforcement learning (RL). We show that the number of samples needed to learn a near-optimal policy with FQI-LOG scales with the accumulated cost of the optimal policy, which is zero in problems where acting optimally achieves the goal and incurs no cost. In doing so, we provide a general framework for proving $\textit{small-cost}$ bounds, i.e. bounds that scale with the optimal achievable cost, in batch RL. Moreover, we empirically verify that FQI-LOG uses fewer samples than FQI trained with squared loss on problems where the optimal policy reliably achieves the goal.
We study Risk-Sensitive Reinforcement Learning (RSRL) with the Optimized Certainty Equivalent (OCE) risk, which generalizes Conditional Value-at-risk (CVaR), entropic risk and Markowitz's mean-variance. Using an augmented Markov Decision Process (MDP), we propose two general meta-algorithms via reductions to standard RL: one based on optimistic algorithms and another based on policy optimization. Our optimistic meta-algorithm generalizes almost all prior RSRL theory with entropic risk or CVaR. Under discrete rewards, our optimistic theory also certifies the first RSRL regret bounds for MDPs with bounded coverability, e.g., exogenous block MDPs. Under discrete rewards, our policy optimization meta-algorithm enjoys both global convergence and local improvement guarantees in a novel metric that lower bounds the true OCE risk. Finally, we instantiate our framework with PPO, construct an MDP, and show that it learns the optimal risk-sensitive policy while prior algorithms provably fail.
Developing accurate off-policy estimators is crucial for both evaluating and optimizing for new policies. The main challenge in off-policy estimation is the distribution shift between the logging policy that generates data and the target policy that we aim to evaluate. Typically, techniques for correcting distribution shift involve some form of importance sampling. This approach results in unbiased value estimation but often comes with the trade-off of high variance, even in the simpler case of one-step contextual bandits. Furthermore, importance sampling relies on the common support assumption, which becomes impractical when the action space is large. To address these challenges, we introduce the Policy Convolution (PC) family of estimators. These methods leverage latent structure within actions -- made available through action embeddings -- to strategically convolve the logging and target policies. This convolution introduces a unique bias-variance trade-off, which can be controlled by adjusting the amount of convolution. Our experiments on synthetic and benchmark datasets demonstrate remarkable mean squared error (MSE) improvements when using PC, especially when either the action space or policy mismatch becomes large, with gains of up to 5 - 6 orders of magnitude over existing estimators.
In this paper, we present empirical studies on conversational recommendation tasks using representative large language models in a zero-shot setting with three primary contributions. (1) Data: To gain insights into model behavior in "in-the-wild" conversational recommendation scenarios, we construct a new dataset of recommendation-related conversations by scraping a popular discussion website. This is the largest public real-world conversational recommendation dataset to date. (2) Evaluation: On the new dataset and two existing conversational recommendation datasets, we observe that even without fine-tuning, large language models can outperform existing fine-tuned conversational recommendation models. (3) Analysis: We propose various probing tasks to investigate the mechanisms behind the remarkable performance of large language models in conversational recommendation. We analyze both the large language models' behaviors and the characteristics of the datasets, providing a holistic understanding of the models' effectiveness, limitations and suggesting directions for the design of future conversational recommenders
Recommender systems aim to answer the following question: given the items that a user has interacted with, what items will this user likely interact with next? Historically this problem is often framed as a predictive task via (self-)supervised learning. In recent years, we have seen more emphasis placed on approaching the recommendation problem from a policy optimization perspective: learning a policy that maximizes some reward function (e.g., user engagement). However, it is commonly the case in recommender systems that we are only able to train a new policy given data collected from a previously-deployed policy. The conventional way to address such a policy mismatch is through importance sampling correction, which unfortunately comes with its own limitations. In this paper, we suggest an alternative approach, which involves the use of local policy improvement without off-policy correction. Drawing from a number of related results in the fields of causal inference, bandits, and reinforcement learning, we present a suite of methods that compute and optimize a lower bound of the expected reward of the target policy. Crucially, this lower bound is a function that is easy to estimate from data, and which does not involve density ratios (such as those appearing in importance sampling correction). We argue that this local policy improvement paradigm is particularly well suited for recommender systems, given that in practice the previously-deployed policy is typically of reasonably high quality, and furthermore it tends to be re-trained frequently and gets continuously updated. We discuss some practical recipes on how to apply some of the proposed techniques in a sequential recommendation setting.
Variational Auto-Encoders (VAEs) have been widely applied for learning compact low-dimensional latent representations for high-dimensional data. When the correlation structure among data points is available, previous work proposed Correlated Variational Auto-Encoders (CVAEs) which employ a structured mixture model as prior and a structured variational posterior for each mixture component to enforce the learned latent representations to follow the same correlation structure. However, as we demonstrate in this paper, such a choice can not guarantee that CVAEs can capture all of the correlations. Furthermore, it prevents us from obtaining a tractable joint and marginal variational distribution. To address these issues, we propose Adaptive Correlated Variational Auto-Encoders (ACVAEs), which apply an adaptive prior distribution that can be adjusted during training, and learn a tractable joint distribution via a saddle-point optimization procedure. Its tractable form also enables further refinement with belief propagation. Experimental results on two real datasets show that ACVAEs outperform other benchmarks significantly.
Variational Auto-Encoders (VAEs) are capable of learning latent representations for high dimensional data. However, due to the i.i.d. assumption, VAEs only optimize the singleton variational distributions and fail to account for the correlations between data points, which might be crucial for learning latent representations from dataset where a priori we know correlations exist. We propose Correlated Variational Auto-Encoders (CVAEs) that can take the correlation structure into consideration when learning latent representations with VAEs. CVAEs apply a prior based on the correlation structure. To address the intractability introduced by the correlated prior, we develop an approximation by average of a set of tractable lower bounds over all maximal acyclic subgraphs of the undirected correlation graph. Experimental results on matching and link prediction on public benchmark rating datasets and spectral clustering on a synthetic dataset show the effectiveness of the proposed method over baseline algorithms.