General-purpose robots require diverse repertoires of behaviors to complete challenging tasks in real-world unstructured environments. To address this issue, goal-conditioned reinforcement learning aims to acquire policies that can reach configurable goals for a wide range of tasks on command. However, such goal-conditioned policies are notoriously difficult and time-consuming to train from scratch. In this paper, we propose Planning to Practice (PTP), a method that makes it practical to train goal-conditioned policies for long-horizon tasks that require multiple distinct types of interactions to solve. Our approach is based on two key ideas. First, we decompose the goal-reaching problem hierarchically, with a high-level planner that sets intermediate subgoals using conditional subgoal generators in the latent space for a low-level model-free policy. Second, we propose a hybrid approach which first pre-trains both the conditional subgoal generator and the policy on previously collected data through offline reinforcement learning, and then fine-tunes the policy via online exploration. This fine-tuning process is itself facilitated by the planned subgoals, which breaks down the original target task into short-horizon goal-reaching tasks that are significantly easier to learn. We conduct experiments in both the simulation and real world, in which the policy is pre-trained on demonstrations of short primitive behaviors and fine-tuned for temporally extended tasks that are unseen in the offline data. Our experimental results show that PTP can generate feasible sequences of subgoals that enable the policy to efficiently solve the target tasks.

In recent years, reinforcement learning (RL) has acquired a prominent position in the space of health-related sequential decision-making, becoming an increasingly popular tool for delivering adaptive interventions (AIs). However, despite potential benefits, its real-life application is still limited, partly due to a poor synergy between the methodological and the applied communities. In this work, we provide the first unified survey on RL methods for learning AIs, using the common methodological umbrella of RL to bridge the two AI areas of dynamic treatment regimes and just-in-time adaptive interventions in mobile health. We outline similarities and differences between these two AI domains and discuss their implications for using RL. Finally, we leverage our experience in designing case studies in both areas to illustrate the tremendous collaboration opportunities between statistical, RL, and healthcare researchers in the space of AIs.

Cost-effective asset management is an area of interest across several industries. Specifically, this paper develops a deep reinforcement learning (DRL) solution to automatically determine an optimal rehabilitation policy for continuously deteriorating water pipes. We approach the problem of rehabilitation planning in an online and offline DRL setting. In online DRL, the agent interacts with a simulated environment of multiple pipes with distinct lengths, materials, and failure rate characteristics. We train the agent using deep Q-learning (DQN) to learn an optimal policy with minimal average costs and reduced failure probability. In offline learning, the agent uses static data, e.g., DQN replay data, to learn an optimal policy via a conservative Q-learning algorithm without further interactions with the environment. We demonstrate that DRL-based policies improve over standard preventive, corrective, and greedy planning alternatives. Additionally, learning from the fixed DQN replay dataset in an offline setting further improves the performance. The results warrant that the existing deterioration profiles of water pipes consisting of large and diverse states and action trajectories provide a valuable avenue to learn rehabilitation policies in the offline setting, which can be further fine-tuned using the simulator.

Evolutionary algorithms have been used to evolve a population of actors to generate diverse experiences for training reinforcement learning agents, which helps to tackle the temporal credit assignment problem and improves the exploration efficiency. However, when adapting this approach to address constrained problems, balancing the trade-off between the reward and constraint violation is hard. In this paper, we propose a novel evolutionary constrained reinforcement learning (ECRL) algorithm, which adaptively balances the reward and constraint violation with stochastic ranking, and at the same time, restricts the policy's behaviour by maintaining a set of Lagrange relaxation coefficients with a constraint buffer. Extensive experiments on robotic control benchmarks show that our ECRL achieves outstanding performance compared to state-of-the-art algorithms. Ablation analysis shows the benefits of introducing stochastic ranking and constraint buffer.

One of the most exciting concepts in the world of artificial intelligence is reinforcement learning. Reinforcement learning is a type of machine learning that involves training an algorithm to make decisions in an environment, with the goal of maximizing a reward. This concept has been applied to a variety of fields, from robotics to game development, and has shown great promise in improving the performance of intelligent systems. At its core, reinforcement learning involves three main components: the environment, the agent, and the reward signal. The environment is the context in which the agent operates and makes decisions.

In the final course from the Machine Learning for Trading specialization, you will be introduced to reinforcement learning (RL) and the benefits of using reinforcement learning in trading strategies. You will learn how RL has been integrated with neural networks and review LSTMs and how they can be applied to time series data. By the end of the course, you will be able to build trading strategies using reinforcement learning, differentiate between actor-based policies and value-based policies, and incorporate RL into a momentum trading strategy. To be successful in this course, you should have advanced competency in Python programming and familiarity with pertinent libraries for machine learning, such as Scikit-Learn, StatsModels, and Pandas. Experience with SQL is recommended.

Data selection is essential for any data-based optimization technique, such as Reinforcement Learning. State-of-the-art sampling strategies for the experience replay buffer improve the performance of the Reinforcement Learning agent. However, they do not incorporate uncertainty in the Q-Value estimation. Consequently, they cannot adapt the sampling strategies, including exploration and exploitation of transitions, to the complexity of the task. To address this, this paper proposes a new sampling strategy that leverages the exploration-exploitation trade-off. This is enabled by the uncertainty estimation of the Q-Value function, which guides the sampling to explore more significant transitions and, thus, learn a more efficient policy. Experiments on classical control environments demonstrate stable results across various environments. They show that the proposed method outperforms state-of-the-art sampling strategies for dense rewards w.r.t. convergence and peak performance by 26% on average.

Energy management systems (EMS) have classically been implemented based on rule-based control (RBC) and model predictive control (MPC) methods. Recent research are investigating reinforcement learning (RL) as a new promising approach. This paper introduces TreeC, a machine learning method that uses the metaheuristic algorithm covariance matrix adaptation evolution strategy (CMA-ES) to generate an interpretable EMS modeled as a decision tree. This method learns the decision strategy of the EMS based on historical data contrary to RBC and MPC approaches that are typically considered as non adaptive solutions. The decision strategy of the EMS is modeled as a decision tree and is thus interpretable contrary to RL which mainly uses black-box models (e.g. neural networks). The TreeC method is compared to RBC, MPC and RL strategies in two study cases taken from literature: (1) an electric grid case and (2) a household heating case. The results show that TreeC obtains close performances than MPC with perfect forecast in both cases and obtains similar performances to RL in the electrical grid case and outperforms RL in the household heating case. TreeC demonstrates a performant application of machine learning for energy management systems that is also fully interpretable.

This work compares two approaches to provide metacognitive interventions and their impact on preparing students for future learning across Intelligent Tutoring Systems (ITSs). In two consecutive semesters, we conducted two classroom experiments: Exp. 1 used a classic artificial intelligence approach to classify students into different metacognitive groups and provide static interventions based on their classified groups. In Exp. 2, we leveraged Deep Reinforcement Learning (DRL) to provide adaptive interventions that consider the dynamic changes in the student's metacognitive levels. In both experiments, students received these interventions that taught how and when to use a backward-chaining (BC) strategy on a logic tutor that supports a default forward-chaining strategy. Six weeks later, we trained students on a probability tutor that only supports BC without interventions. Our results show that adaptive DRL-based interventions closed the metacognitive skills gap between students. In contrast, static classifier-based interventions only benefited a subset of students who knew how to use BC in advance. Additionally, our DRL agent prepared the experimental students for future learning by significantly surpassing their control peers on both ITSs.

We implemented and evaluated an automated cyber defense agent. The agent takes security alerts as input and uses reinforcement learning to learn a policy for executing predefined defensive measures. The defender policies were trained in an environment intended to simulate a cyber attack. In the simulation, an attacking agent attempts to capture targets in the environment, while the defender attempts to protect them by enabling defenses. The environment was modeled using attack graphs based on the Meta Attack Language language. We assumed that defensive measures have downtime costs, meaning that the defender agent was penalized for using them. We also assumed that the environment was equipped with an imperfect intrusion detection system that occasionally produces erroneous alerts based on the environment state. To evaluate the setup, we trained the defensive agent with different volumes of intrusion detection system noise. We also trained agents with different attacker strategies and graph sizes. In experiments, the defensive agent using policies trained with reinforcement learning outperformed agents using heuristic policies. Experiments also demonstrated that the policies could generalize across different attacker strategies. However, the performance of the learned policies decreased as the attack graphs increased in size.