Search for: All records

Meta reinforcement learning (Meta-RL) is an approach wherein the experience gained from solving a variety of tasks is distilled into a meta-policy. The metapolicy, when adapted over only a small (or just a single) number of steps, is able to perform near-optimally on a new, related task. However, a major challenge to adopting this approach to solve real-world problems is that they are often associated with sparse reward functions that only indicate whether a task is completed partially or fully. We consider the situation where some data, possibly generated by a suboptimal agent, is available for each task. We then develop a class of algorithms entitled Enhanced Meta-RL using Demonstrations (EMRLD) that exploit this information—even if sub-optimal—to obtain guidance during training. We show how EMRLD jointly utilizes RL and supervised learning over the offline data to generate a meta-policy that demonstrates monotone performance improvements. We also develop a warm started variant called EMRLD-WS that is particularly efficient for sub-optimal demonstration data. Finally, we show that our EMRLD algorithms significantly outperform existing approaches in a variety of sparse reward environments, including that of a mobile robot. 

Safe reinforcement learning is extremely challenging--not only must the agent explore an unknown environment, it must do so while ensuring no safety constraint violations. We formulate this safe reinforcement learning (RL) problem using the framework of a finite-horizon Constrained Markov Decision Process (CMDP) with an unknown transition probability function, where we model the safety requirements as constraints on the expected cumulative costs that must be satisfied during all episodes of learning. We propose a model-based safe RL algorithm that we call Doubly Optimistic and Pessimistic Exploration (DOPE), and show that it achieves an objective regret $\tilde{O}(|\mathcal{S}|\sqrt{|\mathcal{A}| K})$ without violating the safety constraints during learning, where $|\mathcal{S}|$ is the number of states, $|\mathcal{A}|$ is the number of actions, and $K$ is the number of learning episodes. Our key idea is to combine a reward bonus for exploration (optimism) with a conservative constraint (pessimism), in addition to the standard optimistic model-based exploration. DOPE is not only able to improve the objective regret bound, but also shows a significant empirical performance improvement as compared to earlier optimism-pessimism approaches. 

Crucial performance metrics of a caching algorithm include its ability to quickly and accurately learn a popularity distribution of requests. However, a majority of work on analytical performance analysis focuses on hit probability after an asymptotically large time has elapsed. We consider an online learning viewpoint, and characterize the ``regret'' in terms of the finite time difference between the hits achieved by a candidate caching algorithm with respect to a genie-aided scheme that places the most popular items in the cache. We first consider the Full Observation regime wherein all requests are seen by the cache. We show that the Least Frequently Used (LFU) algorithm is able to achieve order optimal regret, which is matched by an efficient counting algorithm design that we call LFU-Lite. We then consider the Partial Observation regime wherein only requests for items currently cached are seen by the cache, making it similar to an online learning problem related to the multi-armed bandit problem. We show how approaching this ``caching bandit'' using traditional approaches yields either high complexity or regret, but a simple algorithm design that exploits the structure of the distribution can ensure order optimal regret. We conclude by illustrating our insights using numerical simulations. 

The ability of a P2P network to scale its throughput up in proportion to the arrival rate of peers has recently been shown to be crucially dependent on the chunk sharing policy employed. Some policies can result in low frequencies of a particular chunk, known as the missing chunk syndrome, which can dramatically reduce throughput and lead to instability of the system. For instance, commonly used policies that nominally ``boost'' the sharing of infrequent chunks such as the well-known rarest-first algorithm have been shown to be unstable. We take a complementary viewpoint, and instead consider a policy that simply prevents the sharing of the most frequent chunk(s), that we call mode-suppression. We also consider a more general version that suppresses the mode only if the mode frequency is larger than the lowest frequency by a fixed threshold. We prove the stability of mode-suppression using Lyapunov techniques, and use a Kingman bound argument to show that the total download time does not increase with peer arrival rate. We then design versions of mode-suppression that sample a small number of peers at each time, and construct noisy mode estimates by aggregating these samples over time. We show numerically that mode suppression stabilizes and outperforms all other recently proposed chunk sharing algorithms, and via integration into BitTorrent implementation operating over the ns-3 that it ensures stable, low sojourn time operation in a real-world setting. 

Mean Field Games (MFG) are the class of games with a very large number of agents and the standard equilibrium concept is a Mean Field Equilibrium (MFE). Algorithms for learning MFE in dynamic MFGs are unknown in general. Our focus is on an important subclass that possess a monotonicity property called Strategic Complementarities (MFG-SC). We introduce a natural refinement to the equilibrium concept that we call Trembling-Hand-Perfect MFE (T-MFE), which allows agents to employ a measure of randomization while accounting for the impact of such randomization on their payoffs. We propose a simple algorithm for computing T-MFE under a known model. We also introduce a model-free and a model-based approach to learning T-MFE and provide sample complexities of both algorithms. We also develop a fully online learning scheme that obviates the need for a simulator. Finally, we empirically evaluate the performance of the proposed algorithms via examples motivated by real-world applications.