We address the problem of safe reinforcement learning from pixel observations. Inherent challenges in such settings are (1) a trade-off between reward optimization and adhering to safety constraints, (2) partial observability, and (3) high-dimensional observations. We formalize the problem in a constrained, partially observable Markov decision process framework, where an agent obtains distinct reward and safety signals. To address the curse of dimensionality, we employ a novel safety critic using the stochastic latent actor-critic (SLAC) approach. The latent variable model predicts rewards and safety violations, and we use the safety critic to train safe policies. Using well-known benchmark environments, we demonstrate competitive performance over existing approaches with respects to computational requirements, final reward return, and satisfying the safety constraints.

We study a type of Online Linear Programming (OLP) problem that maximizes the objective function with stochastic inputs. The performance of various algorithms that analyze this type of OLP is well studied when the stochastic inputs follow some i.i.d distribution. The two central questions to ask are: (i) can the algorithms achieve the same efficiency if the stochastic inputs are not i.i.d but still stationary, and (ii) how can we modify our algorithms if we know the stochastic inputs are trendy, hence not stationary. We answer the first question by analyzing a regenerative type of input and show the regrets of two popular algorithms are bounded by the same orders as their i.i.d counterparts. We discuss the second question in the context of linearly growing inputs and propose a trend-adaptive algorithm. We provide numerical simulations to illustrate the performance of our algorithms under both regenerative and trendy inputs.

The two most important algorithms in artificial intelligence are backpropagation and belief propagation. In spite of their importance, the connection between them is poorly characterized. We show that when an input to backpropagation is converted into an input to belief propagation so that (loopy) belief propagation can be run on it, then the result of belief propagation encodes the result of backpropagation; thus backpropagation is recovered as a special case of belief propagation. In other words, we prove for apparently the first time that belief propagation generalizes backpropagation. Our analysis is a theoretical contribution, which we motivate with the expectation that it might reconcile our understandings of each of these algorithms, and serve as a guide to engineering researchers seeking to improve the behavior of systems that use one or the other.

As humans, we perceive the three-dimensional structure of the world around us with apparent ease. Think of how vivid the three-dimensional percept is when you look at a vase of flowers sitting on the table next to you. You can tell the shape and translucency of each petal through the subtle patterns of light and shading that play across its surface and effortlessly segment each flower from the background of the scene (Figure 1.1). Looking at a framed group por- trait, you can easily count (and name) all of the people in the picture and even guess at their emotions from their facial appearance. Perceptual psychologists have spent decades trying to understand how the visual system works and, even though they can devise optical illusions1 to tease apart some of its principles (Figure 1.3), a complete solution to this puzzle remains elusive (Marr 1982; Palmer 1999; Livingstone 2008).

We provide a framework for probabilistic reasoning in Vadalog-based Knowledge Graphs (KGs), satisfying the requirements of ontological reasoning: full recursion, powerful existential quantification, expression of inductive definitions. Vadalog is a Knowledge Representation and Reasoning (KRR) language based on Warded Datalog+/-, a logical core language of existential rules, with a good balance between computational complexity and expressive power. Handling uncertainty is essential for reasoning with KGs. Yet Vadalog and Warded Datalog+/- are not covered by the existing probabilistic logic programming and statistical relational learning approaches for several reasons, including insufficient support for recursion with existential quantification, and the impossibility to express inductive definitions. In this work, we introduce Soft Vadalog, a probabilistic extension to Vadalog, satisfying these desiderata. A Soft Vadalog program induces what we call a Probabilistic Knowledge Graph (PKG), which consists of a probability distribution on a network of chase instances, structures obtained by grounding the rules over a database using the chase procedure. We exploit PKGs for probabilistic marginal inference. We discuss the theory and present MCMC-chase, a Monte Carlo method to use Soft Vadalog in practice. We apply our framework to solve data management and industrial problems, and experimentally evaluate it in the Vadalog system.

As an emerging technology in the era of Industry 4.0, digital twin is gaining unprecedented attention because of its promise to further optimize process design, quality control, health monitoring, decision and policy making, and more, by comprehensively modeling the physical world as a group of interconnected digital models. In a two-part series of papers, we examine the fundamental role of different modeling techniques, twinning enabling technologies, and uncertainty quantification and optimization methods commonly used in digital twins. This first paper presents a thorough literature review of digital twin trends across many disciplines currently pursuing this area of research. Then, digital twin modeling and twinning enabling technologies are further analyzed by classifying them into two main categories: physical-to-virtual, and virtual-to-physical, based on the direction in which data flows. Finally, this paper provides perspectives on the trajectory of digital twin technology over the next decade, and introduces a few emerging areas of research which will likely be of great use in future digital twin research. In part two of this review, the role of uncertainty quantification and optimization are discussed, a battery digital twin is demonstrated, and more perspectives on the future of digital twin are shared.

This paper analyzes the simulation to reality gap in reinforcement learning (RL) cyber-physical systems with fractional delays (i.e. delays that are non-integer multiple of the sampling period). The consideration of fractional delay has important implications on the nature of the cyber-physical system considered. Systems with delays are non-Markovian, and the system state vector needs to be extended to make the system Markovian. We show that this is not possible when the delay is in the output, and the problem would always be non-Markovian. Based on this analysis, a sampling scheme is proposed that results in efficient RL training and agents that perform well in realistic multirotor unmanned aerial vehicle simulations. We demonstrate that the resultant agents do not produce excessive oscillations, which is not the case with RL agents that do not consider time delay in the model.

Well-executed emergency evacuations can save lives and reduce suffering. However, decision makers struggle to determine optimal evacuation policies given the chaos, uncertainty, and value judgments inherent in emergency evacuations. We propose and analyze a decision support tool for pre-crisis training exercises for teams preparing for civilian evacuations and explore the tool in the case of the 2021 U.S.-led evacuation from Afghanistan. We use different classes of Markov decision processes (MDPs) to capture compounding levels of uncertainty in (1) the priority category of who appears next at the gate for evacuation, (2) the distribution of priority categories at the population level, and (3) individuals' claimed priority category. We compare the number of people evacuated by priority status under eight heuristic policies. The optimized MDP policy achieves the best performance compared to all heuristic baselines. We also show that accounting for the compounding levels of model uncertainty incurs added complexity without improvement in policy performance. Useful heuristics can be extracted from the optimized policies to inform human decision makers. We open-source all tools to encourage robust dialogue about the trade-offs, limitations, and potential of integrating algorithms into high-stakes humanitarian decision-making.

Training recurrent neural networks is predominantly achieved via backpropagation through time (BPTT). However, this algorithm is not an optimal solution from both a biological and computational perspective. A more efficient and biologically plausible alternative for BPTT is e-prop. We investigate the applicability of e-prop to long short-term memorys (LSTMs), for both supervised and reinforcement learning (RL) tasks. We show that e-prop is a suitable optimization algorithm for LSTMs by comparing it to BPTT on two benchmarks for supervised learning. This proves that e-prop can achieve learning even for problems with long sequences of several hundred timesteps. We introduce extensions that improve the performance of e-prop, which can partially be applied to other network architectures. With the help of these extensions we show that, under certain conditions, e-prop can outperform BPTT for one of the two benchmarks for supervised learning. Finally, we deliver a proof of concept for the integration of e-prop to RL in the domain of deep recurrent Q-learning.

Reinforcement learning (RL) is a decision-making process that seeks to maximize the expected reward when an agent interacts with the environment [Sutton and Barto, 2018]. Over the past decade, RL has gained increasing attention due to its successes in a wide range of domains, including Atari games [Mnih et al., 2013], Go game [Silver et al., 2016], autonomous driving [Yurtsever et al., 2020], Robotics [Kober et al., 2013], etc. Existing RL algorithms can be categorized into value-based algorithms such as Q-learning [Watkins, 1989] and policy-based algorithms such as policy gradient [Sutton et al., 1999]. They can also be categorized as a model-free approach where one directly models the value function classes, or alternatively, a model-based approach where one needs to estimate the transition probability. Due to the intractably large state and action spaces that are used to model the real-world complex environment, function approximation in RL has become prominent in both algorithm design and theoretical analysis. It is a pressing challenge to design sample-efficient RL algorithms with general function approximations. In the special case where the underlying Markov Decision Processes (MDPs) enjoy certain linear structures, several lines of works have achieved polynomial sample complexity and/or T regret guarantees under either model-free or model-based RL settings. For linear MDPs where the transition probability and the reward function admit linear structure, Yang and Wang [2019] developed a variant of Q-learning when granted access to a generative model, Jin et al. [2020] proposed an LSVI-UCB algorithm with a Õ( d