It is challenging to control a soft robot, where reinforcement learning methods have been applied with promising results. However, due to the poor sample efficiency, reinforcement learning methods require a large collection of training data, which limits their applications. In this paper, we propose a Q-learning controller for a physical soft robot, in which pre-trained models using data from a rough simulator are applied to improve the performance of the controller. We implement the method on our soft robot, i.e., Honeycomb Pneumatic Network (HPN) arm. The experiments show that the usage of pre-trained models can not only reduce the amount of the real-world training data, but also greatly improve its accuracy and convergence rate.

Low-resource Relation Extraction (LRE) aims to extract relation facts from limited labeled corpora when human annotation is scarce. Existing works either utilize self-training scheme to generate pseudo labels that will cause the gradual drift problem, or leverage meta-learning scheme which does not solicit feedback explicitly. To alleviate selection bias due to the lack of feedback loops in existing LRE learning paradigms, we developed a Gradient Imitation Reinforcement Learning method to encourage pseudo label data to imitate the gradient descent direction on labeled data and bootstrap its optimization capability through trial and error. We also propose a framework called GradLRE, which handles two major scenarios in low-resource relation extraction. Besides the scenario where unlabeled data is sufficient, GradLRE handles the situation where no unlabeled data is available, by exploiting a contextualized augmentation method to generate data. Experimental results on two public datasets demonstrate the effectiveness of GradLRE on low resource relation extraction when comparing with baselines.

Driven by the need to capture users' evolving interests and optimize their long-term experiences, more and more recommender systems have started to model recommendation as a Markov decision process and employ reinforcement learning to address the problem. Shouldn't research on the fairness of recommender systems follow the same trend from static evaluation and one-shot intervention to dynamic monitoring and non-stop control? In this paper, we portray the recent developments in recommender systems first and then discuss how fairness could be baked into the reinforcement learning techniques for recommendation. Moreover, we argue that in order to make further progress in recommendation fairness, we may want to consider multi-agent (game-theoretic) optimization, multi-objective (Pareto) optimization, and simulation-based optimization, in the general framework of stochastic games.

We propose SLTD (`Sequential Learning-to-Defer') a framework for learning-to-defer pre-emptively to an expert in sequential decision-making settings. SLTD measures the likelihood of improving value of deferring now versus later based on the underlying uncertainty in dynamics. In particular, we focus on the non-stationarity in the dynamics to accurately learn the deferral policy. We demonstrate our pre-emptive deferral can identify regions where the current policy has a low probability of improving outcomes. SLTD outperforms existing non-sequential learning-to-defer baselines, whilst reducing overall uncertainty on multiple synthetic and real-world simulators with non-stationary dynamics. We further derive and decompose the propagated (long-term) uncertainty for interpretation by the domain expert to provide an indication of when the model's performance is reliable.

Early on when learning Reinforcement Learning you may encounter such distinction between algorithms -- some are on-policy some off-policy. You may read many explanations, but still, ask the question: what the hell is the difference? Let's try to clarify this concept once forever. I believe that the best way to do this is by example. So let's set up a simple environment.

Google says it has developed a way of using deep reinforcement learning (RL) to create computer chip floorplanning in just six hours -- a complicated feat that typically requires humans months to achieve. The chips Google's AI develops are on par or superior than those humans can create, the team explained in its paper published in the journal Nature on Wednesday, June 9. In a first for one of its commercial products, Google's research is being used for the company's upcoming tensor processing unit (TPU) chips, which are optimized for AI computation. So Google's AI method to design chips can eventually be used to improve and quicken the future development of AI. "Our method was used to design the next generation of Google's artificial intelligence (AI) accelerators, and has the potential to save thousands of hours of human effort for each new generation," the team said. The major breakthrough is that Google's AI method can be used for chip "floorplanning," which, as the paper said "Despite five decades of research, chip floorplanning has defied automation, requiring months of intense effort by physical design engineers to produce manufacturable layouts."

We explore an online learning reinforcement learning (RL) paradigm for optimizing parallel particle tracing performance in distributed-memory systems. Our method combines three novel components: (1) a workload donation model, (2) a high-order workload estimation model, and (3) a communication cost model, to optimize the performance of data-parallel particle tracing dynamically. First, we design an RL-based workload donation model. Our workload donation model monitors the workload of processes and creates RL agents to donate particles and data blocks from high-workload processes to low-workload processes to minimize the execution time. The agents learn the donation strategy on-the-fly based on reward and cost functions. The reward and cost functions are designed to consider the processes' workload change and the data transfer cost for every donation action. Second, we propose an online workload estimation model, in order to help our RL model estimate the workload distribution of processes in future computations. Third, we design the communication cost model that considers both block and particle data exchange costs, helping the agents make effective decisions with minimized communication cost. We demonstrate that our algorithm adapts to different flow behaviors in large-scale fluid dynamics, ocean, and weather simulation data. Our algorithm improves parallel particle tracing performance in terms of parallel efficiency, load balance, and costs of I/O and communication for evaluations up to 16,384 processors.

Discrete-continuous hybrid action space is a natural setting in many practical problems, such as robot control and game AI. However, most previous Reinforcement Learning (RL) works only demonstrate the success in controlling with either discrete or continuous action space, while seldom take into account the hybrid action space. One naive way to address hybrid action RL is to convert the hybrid action space into a unified homogeneous action space by discretization or continualization, so that conventional RL algorithms can be applied. However, this ignores the underlying structure of hybrid action space and also induces the scalability issue and additional approximation difficulties, thus leading to degenerated results. In this paper, we propose Hybrid Action Representation (HyAR) to learn a compact and decodable latent representation space for the original hybrid action space. HyAR constructs the latent space and embeds the dependence between discrete action and continuous parameter via an embedding table and conditional Variantional Auto-Encoder (VAE). To further improve the effectiveness, the action representation is trained to be semantically smooth through unsupervised environmental dynamics prediction. Finally, the agent then learns its policy with conventional DRL algorithms in the learned representation space and interacts with the environment by decoding the hybrid action embeddings to the original action space. We evaluate HyAR in a variety of environments with discrete-continuous action space. The results demonstrate the superiority of HyAR when compared with previous baselines, especially for high-dimensional action spaces.

We consider the problem of tabular infinite horizon concave utility reinforcement learning (CURL) with convex constraints. Various learning applications with constraints, such as robotics, do not allow for policies that can violate constraints. To this end, we propose a model-based learning algorithm that achieves zero constraint violations. To obtain this result, we assume that the concave objective and the convex constraints have a solution interior to the set of feasible occupation measures. We then solve a tighter optimization problem to ensure that the constraints are never violated despite the imprecise model knowledge and model stochasticity. We also propose a novel Bellman error based analysis for tabular infinite-horizon setups which allows to analyse stochastic policies. Combining the Bellman error based analysis and tighter optimization equation, for $T$ interactions with the environment, we obtain a regret guarantee for objective which grows as $\Tilde{O}(1/\sqrt{T})$, excluding other factors.

Reinforcement learning is a sort of Machine Learning in which an operator learns how and when to respond in a given environment by taking certain actions and observing its outcomes. We can even see a lot of progress in this remarkable field of research in recent decades. DeepMind and the Deep Q learning architecture in 2014, AlphaGo defeating the master of the game of Go in 2016, OpenAI and the PPO in 2017, and others are only a few examples. Reinforcement Learning is based on the premise that an operator can gain knowledge from their environment by interacting with it and obtaining incentives for taking actions. All environmental encounters provide us with opportunities to learn through our interactions with the world.