ترغب بنشر مسار تعليمي؟ اضغط هنا

Diagnosing Bottlenecks in Deep Q-learning Algorithms

65   0   0.0 ( 0 )
 نشر من قبل Justin Fu
 تاريخ النشر 2019
والبحث باللغة English




اسأل ChatGPT حول البحث

Q-learning methods represent a commonly used class of algorithms in reinforcement learning: they are generally efficient and simple, and can be combined readily with function approximators for deep reinforcement learning (RL). However, the behavior of Q-learning methods with function approximation is poorly understood, both theoretically and empirically. In this work, we aim to experimentally investigate potential issues in Q-learning, by means of a unit testing framework where we can utilize oracles to disentangle sources of error. Specifically, we investigate questions related to function approximation, sampling error and nonstationarity, and where available, verify if trends found in oracle settings hold true with modern deep RL methods. We find that large neural network architectures have many benefits with regards to learning stability; offer several practical compensations for overfitting; and develop a novel sampling method based on explicitly compensating for function approximation error that yields fair improvement on high-dimensional continuous control domains.



قيم البحث

اقرأ أيضاً

We investigate the evolution of the Q values for the implementation of Deep Q Learning (DQL) in the Stable Baselines library. Stable Baselines incorporates the latest Reinforcement Learning techniques and achieves superhuman performance in many game environments. However, for some simple non-game environments, the DQL in Stable Baselines can struggle to find the correct actions. In this paper we aim to understand the types of environment where this suboptimal behavior can happen, and also investigate the corresponding evolution of the Q values for individual states. We compare a smart TrafficLight environment (where performance is poor) with the AI Gym FrozenLake environment (where performance is perfect). We observe that DQL struggles with TrafficLight because actions are reversible and hence the Q values in a given state are closer than in FrozenLake. We then investigate the evolution of the Q values using a recent decomposition technique of Achiam et al.. We observe that for TrafficLight, the function approximation error and the complex relationships between the states lead to a situation where some Q values meander far from optimal.
Reinforcement learning (RL) has gained increasing interest since the demonstration it was able to reach human performance on video game benchmarks using deep Q-learning (DQN). The current consensus for training neural networks on such complex environ ments is to rely on gradient-based optimization. Although alternative Bayesian deep learning methods exist, most of them still rely on gradient-based optimization, and they typically do not scale on benchmarks such as the Atari game environment. Moreover none of these approaches allow performing the analytical inference for the weights and biases defining the neural network. In this paper, we present how we can adapt the temporal difference Q-learning framework to make it compatible with the tractable approximate Gaussian inference (TAGI), which allows learning the parameters of a neural network using a closed-form analytical method. Throughout the experiments with on- and off-policy reinforcement learning approaches, we demonstrate that TAGI can reach a performance comparable to backpropagation-trained networks while using fewer hyperparameters, and without relying on gradient-based optimization.
Off-policy learning allows us to learn about possible policies of behavior from experience generated by a different behavior policy. Temporal difference (TD) learning algorithms can become unstable when combined with function approximation and off-po licy sampling - this is known as the deadly triad. Emphatic temporal difference (ETD($lambda$)) algorithm ensures convergence in the linear case by appropriately weighting the TD($lambda$) updates. In this paper, we extend the use of emphatic methods to deep reinforcement learning agents. We show that naively adapting ETD($lambda$) to popular deep reinforcement learning algorithms, which use forward view multi-step returns, results in poor performance. We then derive new emphatic algorithms for use in the context of such algorithms, and we demonstrate that they provide noticeable benefits in small problems designed to highlight the instability of TD methods. Finally, we observed improved performance when applying these algorithms at scale on classic Atari games from the Arcade Learning Environment.
Recent advances in deep reinforcement learning have achieved human-level performance on a variety of real-world applications. However, the current algorithms still suffer from poor gradient estimation with excessive variance, resulting in unstable tr aining and poor sample efficiency. In our paper, we proposed an innovative optimization strategy by utilizing stochastic variance reduced gradient (SVRG) techniques. With extensive experiments on Atari domain, our method outperforms the deep q-learning baselines on 18 out of 20 games.
This paper presents a new neural architecture that combines a modulated Hebbian network (MOHN) with DQN, which we call modulated Hebbian plus Q network architecture (MOHQA). The hypothesis is that such a combination allows MOHQA to solve difficult pa rtially observable Markov decision process (POMDP) problems which impair temporal difference (TD)-based RL algorithms such as DQN, as the TD error cannot be easily derived from observations. The key idea is to use a Hebbian network with bio-inspired neural traces in order to bridge temporal delays between actions and rewards when confounding observations and sparse rewards result in inaccurate TD errors. In MOHQA, DQN learns low level features and control, while the MOHN contributes to the high-level decisions by associating rewards with past states and actions. Thus the proposed architecture combines two modules with significantly different learning algorithms, a Hebbian associative network and a classical DQN pipeline, exploiting the advantages of both. Simulations on a set of POMDPs and on the MALMO environment show that the proposed algorithm improved DQNs results and even outperformed control tests with A2C, QRDQN+LSTM and REINFORCE algorithms on some POMDPs with confounding stimuli and sparse rewards.

الأسئلة المقترحة

التعليقات
جاري جلب التعليقات جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
mircosoft-partner

هل ترغب بارسال اشعارات عن اخر التحديثات في شمرا-اكاديميا