In manipulation tasks like plug insertion or assembly that have low tolerance to errors in pose estimation (errors of the order of 2mm can cause task failure), the utilization of touch/contact modality can aid in accurately localizing the object of interest. Motivated by this, in this work we model high-precision insertion tasks as planning problems under pose uncertainty, where we effectively utilize the occurrence of contacts (or the lack thereof) as observations to reduce uncertainty and reliably complete the task. We present a preprocessing-based planning framework for high-precision insertion in repetitive and time-critical settings, where the set of initial pose distributions (identified by a perception system) is finite. The finite set allows us to enumerate the possible planning problems that can be encountered online and preprocess a database of policies. Due to the computational complexity of constructing this database, we propose a general experience-based POMDP solver, E-RTDP-Bel, that uses the solutions of similar planning problems as experience to speed up planning queries and use it to efficiently construct the database. We show that the developed algorithm speeds up database creation by over a factor of 100, making the process computationally tractable. We demonstrate the effectiveness of the proposed framework in a real-world plug insertion task in the presence of port position uncertainty and a pipe assembly task in simulation in the presence of pipe pose uncertainty.

In unsupervised environment design, reinforcement learning agents are trained on environment configurations (levels) generated by an adversary that maximises some objective. Regret is a commonly used objective that theoretically results in a minimax regret (MMR) policy with desirable robustness guarantees; in particular, the agent's maximum regret is bounded. However, once the agent reaches this regret bound on all levels, the adversary will only sample levels where regret cannot be further reduced. Although there are possible performance improvements to be made outside of these regret-maximising levels, learning stagnates. In this work, we introduce Bayesian level-perfect MMR (BLP), a refinement of the minimax regret objective that overcomes this limitation. We formally show that solving for this objective results in a subset of MMR policies, and that BLP policies act consistently with a Perfect Bayesian policy over all levels. We further introduce an algorithm, ReMiDi, that results in a BLP policy at convergence. We empirically demonstrate that training on levels from a minimax regret adversary causes learning to prematurely stagnate, but that ReMiDi continues learning.

Traditional reinforcement learning control for quadruped robots often relies on privileged information, demanding meticulous selection and precise estimation, thereby imposing constraints on the development process. This work proposes a Self-learning Latent Representation (SLR) method, which achieves high-performance control policy learning without the need for privileged information. To enhance the credibility of our proposed method's evaluation, SLR is compared with open-source code repositories of state-of-the-art algorithms, retaining the original authors' configuration parameters. Across four repositories, SLR consistently outperforms the reference results. Ultimately, the trained policy and encoder empower the quadruped robot to navigate steps, climb stairs, ascend rocks, and traverse various challenging terrains. Robot experiment videos are at https://11chens.github.io/SLR/

Sim-to-real transfer presents a difficult challenge, where models trained in simulation are to be deployed in the real world. The distribution shift between the two settings leads to biased representations of the perceived real-world environment, and thus to suboptimal predictions. In this work, we tackle the challenge of sim-to-real transfer of reinforcement learning (RL) agents for coverage path planning (CPP). In CPP, the task is for a robot to find a path that visits every point of a confined area. Specifically, we consider the case where the environment is unknown, and the agent needs to plan the path online while mapping the environment. We bridge the sim-to-real gap through a semi-virtual environment with a simulated sensor and obstacles, while including real robot kinematics and real-time aspects. We investigate what level of fine-tuning is needed for adapting to a realistic setting, comparing to an agent trained solely in simulation. We find that a high model inference frequency is sufficient for reducing the sim-to-real gap, while fine-tuning degrades performance initially. By training the model in simulation and deploying it at a high inference frequency, we transfer state-of-the-art results from simulation to the real domain, where direct learning would take in the order of weeks with manual interaction, i.e., would be completely infeasible.

Deep Reinforcement Learning (DRL) policies are critically vulnerable to adversarial noise in observations, posing severe risks in safety-critical scenarios. For example, a self-driving car receiving manipulated sensory inputs about traffic signs could lead to catastrophic outcomes. Existing strategies to fortify RL algorithms against such adversarial perturbations generally fall into two categories: (a) using regularization methods that enhance robustness by incorporating adversarial loss terms into the value objectives, and (b) adopting "maximin" principles, which focus on maximizing the minimum value to ensure robustness. While regularization methods reduce the likelihood of successful attacks, their effectiveness drops significantly if an attack does succeed. On the other hand, maximin objectives, although robust, tend to be overly conservative. To address this challenge, we introduce a novel objective called Adversarial Counterfactual Error (ACoE), which naturally balances optimizing value and robustness against adversarial attacks. To optimize ACoE in a scalable manner in model-free settings, we propose a theoretically justified surrogate objective known as Cumulative-ACoE (C-ACoE). The core idea of optimizing C-ACoE is utilizing the belief about the underlying true state given the adversarially perturbed observation. Our empirical evaluations demonstrate that our method outperforms current state-of-the-art approaches for addressing adversarial RL problems across all established benchmarks (MuJoCo, Atari, and Highway) used in the literature.

The problem of reward design examines the interaction between a leader and a follower, where the leader aims to shape the follower's behavior to maximize the leader's payoff by modifying the follower's reward function. Current approaches to reward design rely on an accurate model of how the follower responds to reward modifications, which can be sensitive to modeling inaccuracies. To address this issue of sensitivity, we present a solution that offers robustness against uncertainties in modeling the follower, including 1) how the follower breaks ties in the presence of nonunique best responses, 2) inexact knowledge of how the follower perceives reward modifications, and 3) bounded rationality of the follower. Our robust solution is guaranteed to exist under mild conditions and can be obtained numerically by solving a mixed-integer linear program. Numerical experiments on multiple test cases demonstrate that our solution improves robustness compared to the standard approach without incurring significant additional computing costs.

Temporal distribution shifts are ubiquitous in time series data. One of the most popular methods assumes that the temporal distribution shift occurs uniformly to disentangle the stationary and nonstationary dependencies. But this assumption is difficult to meet, as we do not know when the distribution shifts occur. To solve this problem, we propose to learn IDentifiable latEnt stAtes (IDEA) to detect when the distribution shifts occur. Beyond that, we further disentangle the stationary and nonstationary latent states via sufficient observation assumption to learn how the latent states change. Specifically, we formalize the causal process with environment-irrelated stationary and environment-related nonstationary variables. Under mild conditions, we show that latent environments and stationary/nonstationary variables are identifiable. Based on these theories, we devise the IDEA model, which incorporates an autoregressive hidden Markov model to estimate latent environments and modular prior networks to identify latent states. The IDEA model outperforms several latest nonstationary forecasting methods on various benchmark datasets, highlighting its advantages in real-world scenarios.

The expected regret of any reinforcement learning algorithm is lower bounded by $\Omega\left(\sqrt{DXAT}\right)$ for undiscounted returns, where $D$ is the diameter of the Markov decision process, $X$ the size of the state space, $A$ the size of the action space and $T$ the number of time steps. However, this lower bound is general. A smaller regret can be obtained by taking into account some specific knowledge of the problem structure. In this article, we consider an admission control problem to an $M/M/c/S$ queue with $m$ job classes and class-dependent rewards and holding costs. Queuing systems often have a diameter that is exponential in the buffer size $S$, making the previous lower bound prohibitive for any practical use. We propose an algorithm inspired by UCRL2, and use the structure of the problem to upper bound the expected total regret by $O(S\log T + \sqrt{mT \log T})$ in the finite server case. In the infinite server case, we prove that the dependence of the regret on $S$ disappears.

Markov state modeling has gained popularity in various scientific fields due to its ability to reduce complex time series data into transitions between a few states. Yet, current frameworks are limited by assuming a single Markov chain describes the data, and they suffer an inability to discern heterogeneities. As a solution, this paper proposes a variational expectation-maximization algorithm that identifies a mixture of Markov chains in a time-series data set. The method is agnostic to the definition of the Markov states, whether data-driven (e.g. by spectral clustering) or based on domain knowledge. Variational EM efficiently and organically identifies the number of Markov chains and dynamics of each chain without expensive model comparisons or posterior sampling. The approach is supported by a theoretical analysis and numerical experiments, including simulated and observational data sets based on ${\tt Last.fm}$ music listening, ultramarathon running, and gene expression. The results show the new algorithm is competitive with contemporary mixture modeling approaches and powerful in identifying meaningful heterogeneities in time series data.

Location prediction forecasts a user's location based on historical user mobility traces. To tackle the intrinsic sparsity issue of real-world user mobility traces, spatiotemporal contexts have been shown as significantly useful. Existing solutions mostly incorporate spatiotemporal distances between locations in mobility traces, either by feeding them as additional inputs to Recurrent Neural Networks (RNNs) or by using them to search for informative past hidden states for prediction. However, such distance-based methods fail to capture the time-varying temporal regularities of human mobility, where human mobility is often more regular in the morning than in other periods, for example; this suggests the usefulness of the actual timestamps besides the temporal distances. Against this background, we propose REPLAY, a general RNN architecture learning to capture the time-varying temporal regularities for location prediction. Specifically, REPLAY not only resorts to the spatiotemporal distances in sparse trajectories to search for the informative past hidden states, but also accommodates the time-varying temporal regularities by incorporating smoothed timestamp embeddings using Gaussian weighted averaging with timestamp-specific learnable bandwidths, which can flexibly adapt to the temporal regularities of different strengths across different timestamps. Our extensive evaluation compares REPLAY against a sizable collection of state-of-the-art techniques on two real-world datasets. Results show that REPLAY consistently and significantly outperforms state-of-the-art methods by 7.7\%-10.9\% in the location prediction task, and the bandwidths reveal interesting patterns of the time-varying temporal regularities.