Our team is proposing to run a full-scale energy demand response experiment in an office building. Although this is an exciting endeavor which will provide value to the community, collecting training data for the reinforcement learning agent is costly and will be limited. In this work, we examine how offline training can be leveraged to minimize data costs (accelerate convergence) and program implementation costs. We present two approaches to doing so: pretraining our model to warm start the experiment with simulated tasks, and using a planning model trained to simulate the real world's rewards to the agent. We present results that demonstrate the utility of offline reinforcement learning to efficient price-setting in the energy demand response problem.

In our previous work, we designed a systematic policy to prioritize sampling locations to lead significant accuracy improvement in spatial interpolation by using the prediction uncertainty of Gaussian Process Regression (GPR) as "attraction force" to deployed robots in path planning. Although the integration with Traveling Salesman Problem (TSP) solvers was also shown to produce relatively short travel distance, we here hypothesise several factors that could decrease the overall prediction precision as well because sub-optimal locations may eventually be included in their paths. To address this issue, in this paper, we first explore "local planning" approaches adopting various spatial ranges within which next sampling locations are prioritized to investigate their effects on the prediction performance as well as incurred travel distance. Also, Reinforcement Learning (RL)-based high-level controllers are trained to adaptively produce blended plans from a particular set of local planners to inherit unique strengths from that selection depending on latest prediction states. Our experiments on use cases of temperature monitoring robots demonstrate that the dynamic mixtures of planners can not only generate sophisticated, informative plans that a single planner could not create alone but also ensure significantly reduced travel distances at no cost of prediction reliability without any assist of additional modules for shortest path calculation.

Microscopic epidemic models are powerful tools for government policy makers to predict and simulate epidemic outbreaks, which can capture the impact of individual behaviors on the macroscopic phenomenon. However, existing models only consider simple rule-based individual behaviors, limiting their applicability. This paper proposes a deep-reinforcement-learning-powered microscopic model named Microscopic Pandemic Simulator (MPS). By replacing rule-based agents with rational agents whose behaviors are driven to maximize rewards, the MPS provides a better approximation of real world dynamics. To efficiently simulate with massive amounts of agents in MPS, we propose Scalable Million-Agent DQN (SMADQN). The MPS allows us to efficiently evaluate the impact of different government strategies. This paper first calibrates the MPS against real-world data in Allegheny, US, then demonstratively evaluates two government strategies: information disclosure and quarantine. The results validate the effectiveness of the proposed method. As a broad impact, this paper provides novel insights for the application of DRL in large scale agent-based networks such as economic and social networks.

Reinforcement learning (RL) is well known for requiring large amounts of data in order for RL agents to learn to perform complex tasks. Recent progress in model-based RL allows agents to be much more data-efficient, as it enables them to learn behaviors of visual environments in imagination by leveraging an internal World Model of the environment. Improved sample efficiency can also be achieved by reusing knowledge from previously learned tasks, but transfer learning is still a challenging topic in RL. Parameter-based transfer learning is generally done using an all-or-nothing approach, where the network's parameters are either fully transferred or randomly initialized. In this work we present a simple alternative approach: fractional transfer learning. The idea is to transfer fractions of knowledge, opposed to discarding potentially useful knowledge as is commonly done with random initialization. Using the World Model-based Dreamer algorithm, we identify which type of components this approach is applicable to, and perform experiments in a new multi-source transfer learning setting. The results show that fractional transfer learning often leads to substantially improved performance and faster learning compared to learning from scratch and random initialization.

Loop-invariant synthesis is the basis of every program verification procedure. Due to its undecidability in general, a tool for invariant synthesis necessarily uses heuristics. Despite the common belief that the design of heuristics is vital for the effective performance of a verifier, little work has been performed toward obtaining the optimal heuristics for each invariant-synthesis tool. Instead, developers have hand-tuned the heuristics of tools. This study demonstrates that we can effectively and automatically learn a good heuristic via reinforcement learning for an invariant synthesizer PCSat. Our experiment shows that PCSat combined with the heuristic learned by reinforcement learning outperforms the state-of-the-art solvers for this task. To the best of our knowledge, this is the first work that investigates learning the heuristics of an invariant synthesis tool.

This is the tenth part of "An Outsider's Tour of Reinforcement Learning." Though I've spent the last few posts casting shade at model-free methods for reinforcement learning, I am not blindly against the model-free paradigm. In fact, the most popular methods in core control systems are model free! The most ubiquitous control scheme out there is PID control, and PID has only three parameters. I'd like to use this post to briefly describe PID control, explain how it is closely connected to many of the most popular methods in machine learning, and then turn to explain what PID brings to the table over the model-free methods that drive contemporary RL research.

Deep reinforcement learning (DRL) algorithms have proven effective in robot navigation, especially in unknown environments, through directly mapping perception inputs into robot control commands. Most existing methods adopt uniform execution duration with robots taking commands at fixed intervals. As such, the length of execution duration becomes a crucial parameter to the navigation algorithm. In particular, if the duration is too short, then the navigation policy would be executed at a high frequency, with increased training difficulty and high computational cost. Meanwhile, if the duration is too long, then the policy becomes unable to handle complex situations, like those with crowded obstacles. It is thus tricky to find the "sweet" duration range; some duration values may render a DRL model to fail to find a navigation path. In this paper, we propose to employ adaptive execution duration to overcome this problem. Specifically, we formulate the navigation task as a Semi-Markov Decision Process (SMDP) problem to handle adaptive execution duration. We also improve the distributed proximal policy optimization (DPPO) algorithm and provide its theoretical guarantee for the specified SMDP problem. We evaluate our approach both in the simulator and on an actual robot. The results show that our approach outperforms the other DRL-based method (with fixed execution duration) by 10.3% in terms of the navigation success rate.

In this paper, we consider the problem of multi-agent navigation in partially observable grid environments. This problem is challenging for centralized planning approaches as they, typically, rely on the full knowledge of the environment. We suggest utilizing the reinforcement learning approach when the agents, first, learn the policies that map observations to actions and then follow these policies to reach their goals. To tackle the challenge associated with learning cooperative behavior, i.e. in many cases agents need to yield to each other to accomplish a mission, we use a mixing Q-network that complements learning individual policies. In the experimental evaluation, we show that such approach leads to plausible results and scales well to large number of agents.

Aspect Sentiment Triplet Extraction (ASTE) is the task of extracting triplets of aspect terms, their associated sentiments, and the opinion terms that provide evidence for the expressed sentiments. Previous approaches to ASTE usually simultaneously extract all three components or first identify the aspect and opinion terms, then pair them up to predict their sentiment polarities. In this work, we present a novel paradigm, ASTE-RL, by regarding the aspect and opinion terms as arguments of the expressed sentiment in a hierarchical reinforcement learning (RL) framework. We first focus on sentiments expressed in a sentence, then identify the target aspect and opinion terms for that sentiment. This takes into account the mutual interactions among the triplet's components while improving exploration and sample efficiency. Furthermore, this hierarchical RLsetup enables us to deal with multiple and overlapping triplets. In our experiments, we evaluate our model on existing datasets from laptop and restaurant domains and show that it achieves state-of-the-art performance. The implementation of this work is publicly available at https://github.com/declare-lab/ASTE-RL.

Tim has been selected for his proposal: 'A Fully Probabilistic Theory of Autonomous Decision Making'. Tim's proposal is about developing a fully probabilistic framework for reinforcement learning to provide reliable and mathematically rigorous uncertainty quantification. In contrast to previous approaches, he proposes to treat both the learning process as well as the model components, such an agent's policy, probabilistically. The approach will combine advances in probabilistic inference and modelling with probabilistic reinforcement learning. This will enable autonomous vehicles and machines to'know what they know' as well as what they don't know, and therefore to operate more safely and reliably.