Model-based reinforcement learning (MBRL) is a sample efficient technique to obtain control policies, yet unavoidable modeling errors often lead performance deterioration. The model in MBRL is often solely fitted to reconstruct dynamics, state observations in particular, while the impact of model error on the policy is not captured by the training objective. This leads to a mismatch between the intended goal of MBRL, enabling good policy and value learning, and the target of the loss function employed in practice, future state prediction. Naive intuition would suggest that value-aware model learning would fix this problem and, indeed, several solutions to this objective mismatch problem have been proposed based on theoretical analysis. However, they tend to be inferior in practice to commonly used maximum likelihood (MLE) based approaches. In this paper we propose the Value-gradient weighted Model Learning (VaGraM), a novel method for value-aware model learning which improves the performance of MBRL in challenging settings, such as small model capacity and the presence of distracting state dimensions. We analyze both MLE and value-aware approaches and demonstrate how they fail to account for exploration and the behavior of function approximation when learning value-aware models and highlight the additional goals that must be met to stabilize optimization in the deep learning setting. We verify our analysis by showing that our loss function is able to achieve high returns on the Mujoco benchmark suite while being more robust than maximum likelihood based approaches.

Autonomous driving systems have witnessed a significant development during the past years thanks to the advance in machine learning-enabled sensing and decision-making algorithms. One critical challenge for their massive deployment in the real world is their safety evaluation. Most existing driving systems are still trained and evaluated on naturalistic scenarios collected from daily life or heuristically-generated adversarial ones. However, the large population of cars, in general, leads to an extremely low collision rate, indicating that the safety-critical scenarios are rare in the collected real-world data. Thus, methods to artificially generate scenarios become crucial to measure the risk and reduce the cost. In this survey, we focus on the algorithms of safety-critical scenario generation in autonomous driving. We first provide a comprehensive taxonomy of existing algorithms by dividing them into three categories: data-driven generation, adversarial generation, and knowledge-based generation. Then, we discuss useful tools for scenario generation, including simulation platforms and packages. Finally, we extend our discussion to five main challenges of current works -- fidelity, efficiency, diversity, transferability, controllability -- and research opportunities lighted up by these challenges.

We study the problem of computing an optimal policy of an infinite-horizon discounted constrained Markov decision process (constrained MDP). Despite the popularity of Lagrangian-based policy search methods used in practice, the oscillation of policy iterates in these methods has not been fully understood, bringing out issues such as violation of constraints and sensitivity to hyper-parameters. To fill this gap, we employ the Lagrangian method to cast a constrained MDP into a constrained saddle-point problem in which max/min players correspond to primal/dual variables, respectively, and develop two single-time-scale policy-based primal-dual algorithms with non-asymptotic convergence of their policy iterates to an optimal constrained policy. Specifically, we first propose a regularized policy gradient primal-dual (RPG-PD) method that updates the policy using an entropy-regularized policy gradient, and the dual via a quadratic-regularized gradient ascent, simultaneously. We prove that the policy primal-dual iterates of RPG-PD converge to a regularized saddle point with a sublinear rate, while the policy iterates converge sublinearly to an optimal constrained policy. We further instantiate RPG-PD in large state or action spaces by including function approximation in policy parametrization, and establish similar sublinear last-iterate policy convergence. Second, we propose an optimistic policy gradient primal-dual (OPG-PD) method that employs the optimistic gradient method to update primal/dual variables, simultaneously. We prove that the policy primal-dual iterates of OPG-PD converge to a saddle point that contains an optimal constrained policy, with a linear rate. To the best of our knowledge, this work appears to be the first non-asymptotic policy last-iterate convergence result for single-time-scale algorithms in constrained MDPs.

In this work, we describe a novel method of training an embedding-matching word-level connectionist temporal classification (CTC) automatic speech recognizer (ASR) such that it directly produces word start times and durations, required by many real-world applications, in addition to the transcription. The word timestamps enable the ASR to output word segmentations and word confusion networks without relying on a secondary model or forced alignment process when testing. Our proposed system has similar word segmentation accuracy as a hybrid DNN-HMM (Deep Neural Network-Hidden Markov Model) system, with less than 3ms difference in mean absolute error in word start times on TIMIT data. At the same time, we observed less than 5% relative increase in the word error rate compared to the non-timestamped system when using the same audio training data and nearly identical model size. We also contribute more rigorous analysis of multiple-hypothesis embedding-matching ASR in general.

Cooperative multi-agent reinforcement learning (MARL) for navigation enables agents to cooperate to achieve their navigation goals. Using emergent communication, agents learn a communication protocol to coordinate and share information that is needed to achieve their navigation tasks. In emergent communication, symbols with no pre-specified usage rules are exchanged, in which the meaning and syntax emerge through training. Learning a navigation policy along with a communication protocol in a MARL environment is highly complex due to the huge state space to be explored. To cope with this complexity, this work proposes a novel neural network architecture, for jointly learning an adaptive state space abstraction and a communication protocol among agents participating in navigation tasks. The goal is to come up with an adaptive abstractor that significantly reduces the size of the state space to be explored, without degradation in the policy performance. Simulation results show that the proposed method reaches a better policy, in terms of achievable rewards, resulting in fewer training iterations compared to the case where raw states or fixed state abstraction are used. Moreover, it is shown that a communication protocol emerges during training which enables the agents to learn better policies within fewer training iterations.

Neural networks trained on model simulations enable amortized inference: A pre-trained network can be stored and re-used for Bayesian inference on millions of data sets (von This work proposes "jointly amortized neural Krause et al., 2022). Crucially, most previous neural approaches approximation" (JANA) of intractable likelihood have tackled either SM or SBI in isolation, but little functions and posterior densities arising in attention has been paid to learning both tasks simultaneously. Bayesian surrogate modeling and simulation-based To address this gap, we propose JANA ("Jointly Amortized inference. We train three complementary networks Neural Approximation"), a Bayesian neural framework for in an end-to-end fashion: 1) a summary network simultaneously amortized SM and SBI, and show how it enables to compress individual data points, sets, or time novel solutions to challenging downstream tasks like series into informative embedding vectors; 2) a posterior the estimation of marginal and posterior predictive distributions network to learn an amortized approximate (see Figure 1). JANA also presents a major qualitative posterior; and 3) a likelihood network to learn an upgrade to the BayesFlow framework (Radev et al., 2020), amortized approximate likelihood. Their interaction which was originally designed for amortized SBI alone.

The thesis advances the field of software security by providing knowledge and automation support for software vulnerability assessment using data-driven approaches. Software vulnerability assessment provides important and multifaceted information to prevent and mitigate dangerous cyber-attacks in the wild. The key contributions include a systematisation of knowledge, along with a suite of novel data-driven techniques and practical recommendations for researchers and practitioners in the area. The thesis results help improve the understanding and inform the practice of assessing ever-increasing vulnerabilities in real-world software systems. This in turn enables more thorough and timely fixing prioritisation and planning of these critical security issues.

Autonomous agents deployed in the real world need to be robust against adversarial attacks on sensory inputs. Robustifying agent policies requires anticipating the strongest attacks possible. We demonstrate that existing observation-space attacks on reinforcement learning agents have a common weakness: while effective, their lack of temporal consistency makes them detectable using automated means or human inspection. Detectability is undesirable to adversaries as it may trigger security escalations. We introduce perfect illusory attacks, a novel form of adversarial attack on sequential decision-makers that is both effective and provably statistically undetectable. We then propose the more versatile R-attacks, which result in observation transitions that are consistent with the state-transition function of the adversary-free environment and can be learned end-to-end. Compared to existing attacks, we empirically find R-attacks to be significantly harder to detect with automated methods, and a small study with human subjects suggests they are similarly harder to detect for humans. We propose that undetectability should be a central concern in the study of adversarial attacks on mixed-autonomy settings.

Optimal control of general nonlinear systems is a central challenge in automation. Enabled by powerful function approximators, data-driven approaches to control have recently successfully tackled challenging applications. However, such methods often obscure the structure of dynamics and control behind black-box over-parameterized representations, thus limiting our ability to understand closed-loop behavior. This paper adopts a hybrid-system view of nonlinear modeling and control that lends an explicit hierarchical structure to the problem and breaks down complex dynamics into simpler localized units. We consider a sequence modeling paradigm that captures the temporal structure of the data and derive an expectation-maximization (EM) algorithm that automatically decomposes nonlinear dynamics into stochastic piecewise affine models with nonlinear transition boundaries. Furthermore, we show that these time-series models naturally admit a closed-loop extension that we use to extract local polynomial feedback controllers from nonlinear experts via behavioral cloning. Finally, we introduce a novel hybrid relative entropy policy search (Hb-REPS) technique that incorporates the hierarchical nature of hybrid models and optimizes a set of time-invariant piecewise feedback controllers derived from a piecewise polynomial approximation of a global state-value function.

We consider a system of several collocated nodes sharing a time slotted wireless channel, and seek a MAC (medium access control) that (i) provides low mean delay, (ii) has distributed control (i.e., there is no central scheduler), and (iii) does not require explicit exchange of state information or control signals. The design of such MAC protocols must keep in mind the need for contention access at light traffic, and scheduled access in heavy traffic, leading to the long-standing interest in hybrid, adaptive MACs. Working in the discrete time setting, for the distributed MAC design, we consider a practical information structure where each node has local information and some common information obtained from overhearing. In this setting, "ZMAC" is an existing protocol that is hybrid and adaptive. We approach the problem via two steps (1) We show that it is sufficient for the policy to be "greedy" and "exhaustive". Limiting the policy to this class reduces the problem to obtaining a queue switching policy at queue emptiness instants. (2) Formulating the delay optimal scheduling as a POMDP (partially observed Markov decision process), we show that the optimal switching rule is Stochastic Largest Queue (SLQ). Using this theory as the basis, we then develop a practical distributed scheduler, QZMAC, which is also tunable. We implement QZMAC on standard off-the-shelf TelosB motes and also use simulations to compare QZMAC with the full-knowledge centralized scheduler, and with ZMAC. We use our implementation to study the impact of false detection while overhearing the common information, and the efficiency of QZMAC. Our simulation results show that the mean delay with QZMAC is close that of the full-knowledge centralized scheduler.