Reinforcement learning frameworks have introduced abstractions to implement and execute algorithms at scale. They assume standardized simulator interfaces but are not concerned with identifying suitable task representations. We present Wield, a first-of-its kind system to facilitate task design for practical reinforcement learning. Through software primitives, Wield enables practitioners to decouple system-interface and deployment-specific configuration from state and action design. To guide experimentation, Wield further introduces a novel task design protocol and classification scheme centred around staged randomization to incrementally evaluate model capabilities.

The goal of imitation learning (IL) is to learn a good policy from high-quality demonstrations. However, the quality of demonstrations in reality can be diverse, since it is easier and cheaper to collect demonstrations from a mix of experts and amateurs. IL in such situations can be challenging, especially when the level of demonstrators' expertise is unknown. We propose a new IL method called \underline{v}ariational \underline{i}mitation \underline{l}earning with \underline{d}iverse-quality demonstrations (VILD), where we explicitly model the level of demonstrators' expertise with a probabilistic graphical model and estimate it along with a reward function. We show that a naive approach to estimation is not suitable to large state and action spaces, and fix its issues by using a variational approach which can be easily implemented using existing reinforcement learning methods. Experiments on continuous-control benchmarks demonstrate that VILD outperforms state-of-the-art methods. Our work enables scalable and data-efficient IL under more realistic settings than before.

This paper proposes a novel deep reinforcement learning architecture that was inspired by previous tree structured architectures which were only useable in discrete action spaces. Policy Prediction Network offers a way to improve sample complexity and performance on continuous control problems in exchange for extra computation at training time but at no cost in computation at rollout time. Our approach integrates a mix between model-free and model-based reinforcement learning. Policy Prediction Network is the first to introduce implicit model-based learning to Policy Gradient algorithms for continuous action space and is made possible via the empirically justified clipping scheme. Our experiments are focused on the MuJoCo environments so that they can be compared with similar work done in this area.

Codes and demo are available. This article explores what are states, actions and rewards in reinforcement learning, and how agent can learn through simulation to determine the best actions to take in any given state. After a long day at work, you are deciding between 2 choices: to head home and write an article or hang out with friends at a bar. If you choose to hang out with friends, your friends will make you feel happy; whereas heading home to write an article, you'll end up feeling tired after a long day at work. In this example, enjoying yourself is a reward and feeling tired is viewed as a negative reward, so why write articles?

Typically, a RL setup is composed of two components, an agent and an environment. Then environment refers to the object that the agent is acting on (e.g. the game itself in the Atari game), while the agent represents the RL algorithm. The environment starts by sending a state to the agent, which then based on its knowledge to take an action in response to that state. After that, the environment send a pair of next state and reward back to the agent. The agent will update its knowledge with the reward returned by the environment to evaluate its last action.

A6 months old baby won't even notice if a toy truck drives off a platform and seems to fly in the air. However, if the same experiment is repeated 2 to 3 months later, the baby will immediately identify that something is wrong. This means that the baby has already learned the concept of gravity. "Nobody tells a baby that objects are supposed to fall," said the chief AI scientist at Facebook and a professor at NYU, Dr. Yann LeCun, during a webinar organized by the Association for Computing Machinery, an industry body. Because babies do not have very sophisticated motor control, LeCun hypothesizes, "a lot of what they learn about the world is through observation."

To operate successfully in a complex and changing environment, learning agents must be able to acquire new skills quickly. Humans display remarkable skill in this area -- we can learn to recognize a new object from one example, adapt to driving a different car in a matter of minutes, and add a new slang word to our vocabulary after hearing it once. Meta-learning is a promising approach for enabling such capabilities in machines. In this paradigm, the agent adapts to a new task from limited data by leveraging a wealth of experience collected in performing related tasks. For agents that must take actions and collect their own experience, meta-reinforcement learning (meta-RL) holds the promise of enabling fast adaptation to new scenarios.

Then environment refers to the object that the agent is acting on (e.g. the game itself in the Atari game), while the agent represents the RL algorithm. The environment starts by sending a state to the agent, which then based on its knowledge to take an action in response to that state. After that, the environment send a pair of next state and reward back to the agent. The agent will update its knowledge with the reward returned by the environment to evaluate its last action. The loop keeps going on until the environment sends a terminal state, which ends to episode.

Real-world graph applications, such as advertisements and product recommendations make profits based on accurately classify the label of the nodes. However, in such scenarios, there are high incentives for the adversaries to attack such graph to reduce the node classification performance. Previous work on graph adversarial attacks focus on modifying existing graph structures, which is infeasible in most real-world applications. In contrast, it is more practical to inject adversarial nodes into existing graphs, which can also potentially reduce the performance of the classifier. In this paper, we study the novel node injection poisoning attacks problem which aims to poison the graph. We describe a reinforcement learning based method, namely NIPA, to sequentially modify the adversarial information of the injected nodes. We report the results of experiments using several benchmark data sets that show the superior performance of the proposed method NIPA, relative to the existing state-of-the-art methods.

Traditional planning and control methods could fail to find a feasible trajectory for an autonomous vehicle to execute amongst dense traffic on roads. This is because the obstacle-free volume in spacetime is very small in these scenarios for the vehicle to drive through. However, that does not mean the task is infeasible since human drivers are known to be able to drive amongst dense traffic by leveraging the cooperativeness of other drivers to open a gap. The traditional methods fail to take into account the fact that the actions taken by an agent affect the behaviour of other vehicles on the road. In this work, we rely on the ability of deep reinforcement learning to implicitly model such interactions and learn a continuous control policy over the action space of an autonomous vehicle. The application we consider requires our agent to negotiate and open a gap in the road in order to successfully merge or change lanes. Our policy learns to repeatedly probe into the target road lane while trying to find a safe spot to move in to. We compare against two model-predictive control-based algorithms and show that our policy outperforms them in simulation.