PyTorchKR Papers

최신 논문

총 1,872편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2204.10349

Provably Efficient Kernelized Q-Learning

We propose and analyze a kernelized version of Q-learning. Although a kernel space is typically infinite-dimensional, extensive study has shown that generalization is only affected by the effective dimension of the data. We incorporate such ideas into the Q-learning framework and derive regret bounds for arbitrary kernels. In particular, we provide concrete bounds for linear kernels and Gaussian RBF kernels; notably, the latter bound looks almost identical to the former, only that the actual dimension is replaced by a different notion of dimensionality. Finally, we test our algorithm on a suite of classic control tasks; remarkably, under the Gaussian RBF kernel, it achieves reasonably good performance after only 1000 environmental steps, while its neural network counterpart, deep Q-learning, still struggles.

arXiv:2204.10179

Distributed Learning for Vehicular Dynamic Spectrum Access in Autonomous Driving

Reliable wireless communication between the autonomously driving cars is one of the fundamental needs for guaranteeing passenger safety and comfort. However, when the number of communicating cars increases, the transmission quality may be significantly degraded due to too high occupancy radio of the used frequency band. In this paper, we concentrate on the autonomous vehicle-platooning use-case, where intra-platoon communication is done in the dynamically selected frequency band, other than nominally devoted for such purposes. The carrier selection is done in a flexible manner with the support of the context database located at the roadside unit (edge of wireless communication infrastructure). However, as the database delivers only context information to the platoons' leaders, the final decision is made separately by the individual platoons, following the suggestions made by the artificial intelligence algorithms. In this work, we concentrate on a lightweight Q-learning solution, that could be successfully implemented in each car for dynamic channel selection.

arXiv:2204.06904

Efficient and practical quantum compiler towards multi-qubit systems with deep reinforcement learning

Efficient quantum compiling tactics greatly enhance the capability of quantum computers to execute complicated quantum algorithms. Due to its fundamental importance, a plethora of quantum compilers has been designed in past years. However, there are several caveats to current protocols, which are low optimality, high inference time, limited scalability, and lack of universality. To compensate for these defects, here we devise an efficient and practical quantum compiler assisted by advanced deep reinforcement learning (RL) techniques, i.e., data generation, deep Q-learning, and AQ* search. In this way, our protocol is compatible with various quantum machines and can be used to compile multi-qubit operators. We systematically evaluate the performance of our proposal in compiling quantum operators with both inverse-closed and inverse-free universal basis sets. In the task of single-qubit operator compiling, our proposal outperforms other RL-based quantum compilers in the measure of compiling sequence length and inference time. Meanwhile, the output solution is near-optimal, guaranteed by the Solovay-Kitaev theorem. Notably, for the inverse-free universal basis set, the achieved sequence length complexity is comparable with the inverse-based setting and dramatically advances previous methods. These empirical results contribute to improving the inverse-free Solovay-Kitaev theorem. In addition, for the first time, we demonstrate how to leverage RL-based quantum compilers to accomplish two-qubit operator compiling. The achieved results open an avenue for integrating RL with quantum compiling to unify efficiency and practicality and thus facilitate the exploration of quantum advantages.

arXiv:2103.05893

A Jointly Optimal Design of Control and Scheduling in Networked Systems under Denial-of-Service Attacks

We consider the joint design of control and scheduling under stochastic Denial-of-Service (DoS) attacks in the context of networked control systems. A sensor takes measurements of the system output and forwards its dynamic state estimates to a remote controller over a packet-dropping link. The controller determines the optimal control law for the process using the estimates it receives. An attacker aims at degrading the control performance by increasing the packet-dropout rate with a DoS attack towards the sensor-controller channel. We assume both the controller and the attacker are rational in a game-theoretic sense and establish a partially observable stochastic game to derive the optimal joint design of scheduling and control. Using dynamic programming we prove that the control and scheduling policies can be designed separately without sacrificing optimality, making the problem equivalent to a complete information game. We employ Nash Q-learning to solve the problem and prove that the solution is guaranteed to constitute an $\epsilon$-Nash equilibrium. Numerical examples are provided to illustrate the tradeoffs between control performance and communication cost.

arXiv:2204.03771

Q-learning with online random forests

$Q$-learning is the most fundamental model-free reinforcement learning algorithm. Deployment of $Q$-learning requires approximation of the state-action value function (also known as the $Q$-function). In this work, we provide online random forests as $Q$-function approximators and propose a novel method wherein the random forest is grown as learning proceeds (through expanding forests). We demonstrate improved performance of our methods over state-of-the-art Deep $Q$-Networks in two OpenAI gyms (`blackjack' and `inverted pendulum') but not in the `lunar lander' gym. We suspect that the resilience to overfitting enjoyed by random forests recommends our method for common tasks that do not require a strong representation of the problem domain. We show that expanding forests (in which the number of trees increases as data comes in) improve performance, suggesting that expanding forests are viable for other applications of online random forests beyond the reinforcement learning setting.

arXiv:2204.03487

Optimizing the Long-Term Behaviour of Deep Reinforcement Learning for Pushing and Grasping

We investigate the "Visual Pushing for Grasping" (VPG) system by Zeng et al. and the "Hourglass" system by Ewerton et al., an evolution of the former. The focus of our work is the investigation of the capabilities of both systems to learn long-term rewards and policies. Zeng et al. original task only needs a limited amount of foresight. Ewerton et al. attain their best performance using an agent which only takes the most immediate action under consideration. We are interested in the ability of their models and training algorithms to accurately predict long-term Q-Values. To evaluate this ability, we design a new bin sorting task and reward function. Our task requires agents to accurately estimate future rewards and therefore use high discount factors in their Q-Value calculation. We investigate the behaviour of an adaptation of the VPG training algorithm on our task. We show that this adaptation can not accurately predict the required long-term action sequences. In addition to the limitations identified by Ewerton et al., it suffers from the known Deep Q-Learning problem of overestimated Q-Values. In an effort to solve our task, we turn to the Hourglass models and combine them with the Double Q-Learning approach. We show that this approach enables the models to accurately predict long-term action sequences when trained with large discount factors. Our results show that the Double Q-Learning technique is essential for training with very high discount factors, as the models Q-Value predictions diverge otherwise. We also experiment with different approaches for discount factor scheduling, loss calculation and exploration procedures. Our results show that the latter factors do not visibly influence the model's performance for our task.

arXiv:2204.01844

Deep Q-learning of global optimizer of multiply model parameters for viscoelastic imaging

Objective: Estimation of the global optima of multiple model parameters is valuable in imaging to form a reliable diagnostic image. Given non convexity of the objective function, it is challenging to avoid from different local minima. Methods: We first formulate the global searching of multiply parameters to be a k-D move in the parametric space, and convert parameters updating to be state-action decision-making problem. We proposed a novel Deep Q-learning of Model Parameters (DQMP) method for global optimization of model parameters by updating the parameter configurations through actions that maximize a Q-value, which employs a Deep Reward Network designed to learn global reward values from both visible curve fitting errors and hidden parameter errors. Results: The DQMP method was evaluated by viscoelastic imaging on soft matter by Kelvin-Voigt fractional derivative (KVFD) modeling. In comparison to other methods, imaging of parameters by DQMP yielded the smallest errors (< 2%) to the ground truth images. DQMP was applied to viscoelastic imaging on biological tissues, which indicated a great potential of imaging on physical parameters in diagnostic applications. Conclusions: DQMP method is able to achieve global optima, yielding accurate model parameter estimates in viscoelastic imaging. Assessment of DQMP by simulation imaging and ultrasound breast imaging demonstrated the consistency, reliability of the imaged parameters, and powerful global searching ability of DQMP. Significance: DQMP method is promising for imaging of multiple parameters, and can be generalized to global optimization for many other complex nonconvex functions and imaging of physical parameters.

arXiv:2204.01975공식 코드 · Papers with Code

GAIL-PT: A Generic Intelligent Penetration Testing Framework with Generative Adversarial Imitation Learning

Penetration testing (PT) is an efficient network testing and vulnerability mining tool by simulating a hacker's attack for valuable information applied in some areas. Compared with manual PT, intelligent PT has become a dominating mainstream due to less time-consuming and lower labor costs. Unfortunately, RL-based PT is still challenged in real exploitation scenarios because the agent's action space is usually high-dimensional discrete, thus leading to algorithm convergence difficulty. Besides, most PT methods still rely on the decisions of security experts. Addressing the challenges, for the first time, we introduce expert knowledge to guide the agent to make better decisions in RL-based PT and propose a Generative Adversarial Imitation Learning-based generic intelligent Penetration testing framework, denoted as GAIL-PT, to solve the problems of higher labor costs due to the involvement of security experts and high-dimensional discrete action space. Specifically, first, we manually collect the state-action pairs to construct an expert knowledge base when the pre-trained RL / DRL model executes successful penetration testings. Second, we input the expert knowledge and the state-action pairs generated online by the different RL / DRL models into the discriminator of GAIL for training. At last, we apply the output reward of the discriminator to guide the agent to perform the action with a higher penetration success rate to improve PT's performance. Extensive experiments conducted on the real target host and simulated network scenarios show that GAIL-PT achieves the SOTA penetration performance against DeepExploit in exploiting actual target Metasploitable2 and Q-learning in optimizing penetration path, not only in small-scale with or without honey-pot network environments but also in the large-scale virtual network environment.

arXiv:2203.16989

Functional Stability of Discounted Markov Decision Processes Using Economic MPC Dissipativity Theory

This paper discusses the functional stability of closed-loop Markov Chains under optimal policies resulting from a discounted optimality criterion, forming Markov Decision Processes (MDPs). We investigate the stability of MDPs in the sense of probability measures (densities) underlying the state distributions and extend the dissipativity theory of Economic Model Predictive Control in order to characterize the MDP stability. This theory requires a so-called storage function satisfying a dissipativity inequality. In the probability measures space and for the discounted setting, we introduce new dissipativity conditions ensuring the MDP stability. We then use finite-horizon optimal control problems in order to generate valid storage functionals. In practice, we propose to use Q-learning to compute the storage functionals.

arXiv:2203.17128

Neural Q-learning for solving PDEs

Solving high-dimensional partial differential equations (PDEs) is a major challenge in scientific computing. We develop a new numerical method for solving elliptic-type PDEs by adapting the Q-learning algorithm in reinforcement learning. Our "Q-PDE" algorithm is mesh-free and therefore has the potential to overcome the curse of dimensionality. Using a neural tangent kernel (NTK) approach, we prove that the neural network approximator for the PDE solution, trained with the Q-PDE algorithm, converges to the trajectory of an infinite-dimensional ordinary differential equation (ODE) as the number of hidden units $\rightarrow \infty$. For monotone PDE (i.e. those given by monotone operators, which may be nonlinear), despite the lack of a spectral gap in the NTK, we then prove that the limit neural network, which satisfies the infinite-dimensional ODE, converges in $L^2$ to the PDE solution as the training time $\rightarrow \infty$. More generally, we can prove that any fixed point of the wide-network limit for the Q-PDE algorithm is a solution of the PDE (not necessarily under the monotone condition). The numerical performance of the Q-PDE algorithm is studied for several elliptic PDEs.

arXiv:2203.15955

Investigating the Properties of Neural Network Representations in Reinforcement Learning

In this paper we investigate the properties of representations learned by deep reinforcement learning systems. Much of the early work on representations for reinforcement learning focused on designing fixed-basis architectures to achieve properties thought to be desirable, such as orthogonality and sparsity. In contrast, the idea behind deep reinforcement learning methods is that the agent designer should not encode representational properties, but rather that the data stream should determine the properties of the representation -- good representations emerge under appropriate training schemes. In this paper we bring these two perspectives together, empirically investigating the properties of representations that support transfer in reinforcement learning. We introduce and measure six representational properties over more than 25 thousand agent-task settings. We consider Deep Q-learning agents with different auxiliary losses in a pixel-based navigation environment, with source and transfer tasks corresponding to different goal locations. We develop a method to better understand why some representations work better for transfer, through a systematic approach varying task similarity and measuring and correlating representation properties with transfer performance. We demonstrate the generality of the methodology by investigating representations learned by a Rainbow agent that successfully transfer across games modes in Atari 2600.

arXiv:2203.15845공식 코드 · Papers with Code

Topological Experience Replay

State-of-the-art deep Q-learning methods update Q-values using state transition tuples sampled from the experience replay buffer. This strategy often uniformly and randomly samples or prioritizes data sampling based on measures such as the temporal difference (TD) error. Such sampling strategies can be inefficient at learning Q-function because a state's Q-value depends on the Q-value of successor states. If the data sampling strategy ignores the precision of the Q-value estimate of the next state, it can lead to useless and often incorrect updates to the Q-values. To mitigate this issue, we organize the agent's experience into a graph that explicitly tracks the dependency between Q-values of states. Each edge in the graph represents a transition between two states by executing a single action. We perform value backups via a breadth-first search starting from that expands vertices in the graph starting from the set of terminal states and successively moving backward. We empirically show that our method is substantially more data-efficient than several baselines on a diverse range of goal-reaching tasks. Notably, the proposed method also outperforms baselines that consume more batches of training experience and operates from high-dimensional observational data such as images.

arXiv:2203.13884

A Conservative Q-Learning approach for handling distribution shift in sepsis treatment strategies

Sepsis is a leading cause of mortality and its treatment is very expensive. Sepsis treatment is also very challenging because there is no consensus on what interventions work best and different patients respond very differently to the same treatment. Deep Reinforcement Learning methods can be used to come up with optimal policies for treatment strategies mirroring physician actions. In the healthcare scenario, the available data is mostly collected offline with no interaction with the environment, which necessitates the use of offline RL techniques. The Offline RL paradigm suffers from action distribution shifts which in turn negatively affects learning an optimal policy for the treatment. In this work, a Conservative-Q Learning (CQL) algorithm is used to mitigate this shift and its corresponding policy reaches closer to the physicians policy than conventional deep Q Learning. The policy learned could help clinicians in Intensive Care Units to make better decisions while treating septic patients and improve survival rate.

arXiv:2203.13865공식 코드 · Papers with Code

Intelligent Masking: Deep Q-Learning for Context Encoding in Medical Image Analysis

The need for a large amount of labeled data in the supervised setting has led recent studies to utilize self-supervised learning to pre-train deep neural networks using unlabeled data. Many self-supervised training strategies have been investigated especially for medical datasets to leverage the information available in the much fewer unlabeled data. One of the fundamental strategies in image-based self-supervision is context prediction. In this approach, a model is trained to reconstruct the contents of an arbitrary missing region of an image based on its surroundings. However, the existing methods adopt a random and blind masking approach by focusing uniformly on all regions of the images. This approach results in a lot of unnecessary network updates that cause the model to forget the rich extracted features. In this work, we develop a novel self-supervised approach that occludes targeted regions to improve the pre-training procedure. To this end, we propose a reinforcement learning-based agent which learns to intelligently mask input images through deep Q-learning. We show that training the agent against the prediction model can significantly improve the semantic features extracted for downstream classification tasks. We perform our experiments on two public datasets for diagnosing breast cancer in the ultrasound images and detecting lower-grade glioma with MR images. In our experiments, we show that our novel masking strategy advances the learned features according to the performance on the classification task in terms of accuracy, macro F1, and AUROC.

arXiv:2203.12315

The state-of-the-art review on resource allocation problem using artificial intelligence methods on various computing paradigms

With the increasing growth of information through smart devices, increasing the quality level of human life requires various computational paradigms presentation including the Internet of Things, fog, and cloud. Between these three paradigms, the cloud computing paradigm as an emerging technology adds cloud layer services to the edge of the network so that resource allocation operations occur close to the end-user to reduce resource processing time and network traffic overhead. Hence, the resource allocation problem for its providers in terms of presenting a suitable platform, by using computational paradigms is considered a challenge. In general, resource allocation approaches are divided into two methods, including auction-based methods(goal, increase profits for service providers-increase user satisfaction and usability) and optimization-based methods(energy, cost, network exploitation, Runtime, reduction of time delay). In this paper, according to the latest scientific achievements, a comprehensive literature study (CLS) on artificial intelligence methods based on resource allocation optimization without considering auction-based methods in various computing environments are provided such as cloud computing, Vehicular Fog Computing, wireless, IoT, vehicular networks, 5G networks, vehicular cloud architecture,machine-to-machine communication(M2M),Train-to-Train(T2T) communication network, Peer-to-Peer(P2P) network. Since deep learning methods based on artificial intelligence are used as the most important methods in resource allocation problems; Therefore, in this paper, resource allocation approaches based on deep learning are also used in the mentioned computational environments such as deep reinforcement learning, Q-learning technique, reinforcement learning, online learning, and also Classical learning methods such as Bayesian learning, Cummins clustering, Markov decision process.

arXiv:2203.11489

A Note on Target Q-learning For Solving Finite MDPs with A Generative Oracle

Q-learning with function approximation could diverge in the off-policy setting and the target network is a powerful technique to address this issue. In this manuscript, we examine the sample complexity of the associated target Q-learning algorithm in the tabular case with a generative oracle. We point out a misleading claim in [Lee and He, 2020] and establish a tight analysis. In particular, we demonstrate that the sample complexity of the target Q-learning algorithm in [Lee and He, 2020] is $\widetilde{\mathcal O}(|\mathcal S|^2|\mathcal A|^2 (1-\gamma)^{-5}\varepsilon^{-2})$. Furthermore, we show that this sample complexity is improved to $\widetilde{\mathcal O}(|\mathcal S||\mathcal A| (1-\gamma)^{-5}\varepsilon^{-2})$ if we can sequentially update all state-action pairs and $\widetilde{\mathcal O}(|\mathcal S||\mathcal A| (1-\gamma)^{-4}\varepsilon^{-2})$ if $\gamma$ is further in $(1/2, 1)$. Compared with the vanilla Q-learning, our results conclude that the introduction of a periodically-frozen target Q-function does not sacrifice the sample complexity.

arXiv:2203.11526공식 코드 · Papers with Code

Action Candidate Driven Clipped Double Q-learning for Discrete and Continuous Action Tasks

Double Q-learning is a popular reinforcement learning algorithm in Markov decision process (MDP) problems. Clipped Double Q-learning, as an effective variant of Double Q-learning, employs the clipped double estimator to approximate the maximum expected action value. Due to the underestimation bias of the clipped double estimator, the performance of clipped Double Q-learning may be degraded in some stochastic environments. In this paper, in order to reduce the underestimation bias, we propose an action candidate-based clipped double estimator for Double Q-learning. Specifically, we first select a set of elite action candidates with high action values from one set of estimators. Then, among these candidates, we choose the highest valued action from the other set of estimators. Finally, we use the maximum value in the second set of estimators to clip the action value of the chosen action in the first set of estimators and the clipped value is used for approximating the maximum expected action value. Theoretically, the underestimation bias in our clipped Double Q-learning decays monotonically as the number of action candidates decreases. Moreover, the number of action candidates controls the trade-off between the overestimation and underestimation biases. In addition, we also extend our clipped Double Q-learning to continuous action tasks via approximating the elite continuous action candidates. We empirically verify that our algorithm can more accurately estimate the maximum expected action value on some toy environments and yield good performance on several benchmark problems.

arXiv:2203.10142

Infinite-Horizon Reach-Avoid Zero-Sum Games via Deep Reinforcement Learning

In this paper, we consider the infinite-horizon reach-avoid zero-sum game problem, where the goal is to find a set in the state space, referred to as the reach-avoid set, such that the system starting at a state therein could be controlled to reach a given target set without violating constraints under the worst-case disturbance. We address this problem by designing a new value function with a contracting Bellman backup, where the super-zero level set, i.e., the set of states where the value function is evaluated to be non-negative, recovers the reach-avoid set. Building upon this, we prove that the proposed method can be adapted to compute the viability kernel, or the set of states which could be controlled to satisfy given constraints, and the backward reachable set, or the set of states that could be driven towards a given target set. Finally, we propose to alleviate the curse of dimensionality issue in high-dimensional problems by extending Conservative Q-Learning, a deep reinforcement learning technique, to learn a value function such that the super-zero level set of the learned value function serves as a (conservative) approximation to the reach-avoid set. Our theoretical and empirical results suggest that the proposed method could learn reliably the reach-avoid set and the optimal control policy even with neural network approximation.

arXiv:2203.07500

Reinforcement Learning for Optimal Control of a District Cooling Energy Plant

District cooling energy plants (DCEPs) consisting of chillers, cooling towers, and thermal energy storage (TES) systems consume a considerable amount of electricity. Optimizing the scheduling of the TES and chillers to take advantage of time-varying electricity price is a challenging optimal control problem. The classical method, model predictive control (MPC), requires solving a high dimensional mixed-integer nonlinear program (MINLP) because of the on/off actuation of the chillers and charging/discharging of TES, which are computationally challenging. RL is an attractive alternative to MPC: the real time control computation is a low-dimensional optimization problem that can be easily solved. However, the performance of an RL controller depends on many design choices. In this paper, we propose a Q-learning based reinforcement learning (RL) controller for this problem. Numerical simulation results show that the proposed RL controller is able to reduce energy cost over a rule-based baseline controller by approximately 8%, comparable to savings reported in the literature with MPC for similar DCEPs. We describe the design choices in the RL controller, including basis functions, reward function shaping, and learning algorithm parameters. Compared to existing work on RL for DCEPs, the proposed controller is designed for continuous state and actions spaces.

arXiv:2203.07368

The Efficacy of Pessimism in Asynchronous Q-Learning

This paper is concerned with the asynchronous form of Q-learning, which applies a stochastic approximation scheme to Markovian data samples. Motivated by the recent advances in offline reinforcement learning, we develop an algorithmic framework that incorporates the principle of pessimism into asynchronous Q-learning, which penalizes infrequently-visited state-action pairs based on suitable lower confidence bounds (LCBs). This framework leads to, among other things, improved sample efficiency and enhanced adaptivity in the presence of near-expert data. Our approach permits the observed data in some important scenarios to cover only partial state-action space, which is in stark contrast to prior theory that requires uniform coverage of all state-action pairs. When coupled with the idea of variance reduction, asynchronous Q-learning with LCB penalization achieves near-optimal sample complexity, provided that the target accuracy level is small enough. In comparison, prior works were suboptimal in terms of the dependency on the effective horizon even when i.i.d. sampling is permitted. Our results deliver the first theoretical support for the use of pessimism principle in the presence of Markovian non-i.i.d. data.