arXiv subjects

Xiangfeng Wang

Publications and source records attributed to Xiangfeng Wang.

At least 73 records · Page 4Linked to original sources

arXiv2022

Learning Cooperative Oversubscription for Cloud by Chance-Constrained Multi-Agent Reinforcement Learning

Oversubscription is a common practice for improving cloud resource utilization. It allows the cloud service provider to sell more resources than the physical limit, assuming not all users would fully utilize the resources simultaneously. However, how to design an oversubscription policy that improves utilization while satisfying the some safety constraints remains an open problem. Existing methods and industrial practices are over-conservative, ignoring the coordination of diverse resource usage patterns and probabilistic constraints. To address these two limitations, this paper formulates the oversubscription for cloud as a chance-constrained optimization problem and propose an effective Chance Constrained Multi-Agent Reinforcement Learning (C2MARL) method to solve this problem. Specifically, C2MARL reduces the number of constraints by considering their upper bounds and leverages a multi-agent reinforcement learning paradigm to learn a safe and optimal coordination policy. We evaluate our C2MARL on an internal cloud platform and public cloud datasets. Experiments show that our C2MARL outperforms existing methods in improving utilization ($20\%\sim 86\%$) under different levels of safety constraints.

cs.LG↗

arXiv2022

Dealing with Non-Stationarity in MARL via Trust-Region Decomposition

Non-stationarity is one thorny issue in cooperative multi-agent reinforcement learning (MARL). One of the reasons is the policy changes of agents during the learning process. Some existing works have discussed various consequences caused by non-stationarity with several kinds of measurement indicators. This makes the objectives or goals of existing algorithms are inevitably inconsistent and disparate. In this paper, we introduce a novel notion, the $δ$-measurement, to explicitly measure the non-stationarity of a policy sequence, which can be further proved to be bounded by the KL-divergence of consecutive joint policies. A straightforward but highly non-trivial way is to control the joint policies' divergence, which is difficult to estimate accurately by imposing the trust-region constraint on the joint policy. Although it has lower computational complexity to decompose the joint policy and impose trust-region constraints on the factorized policies, simple policy factorization like mean-field approximation will lead to more considerable policy divergence, which can be considered as the trust-region decomposition dilemma. We model the joint policy as a pairwise Markov random field and propose a trust-region decomposition network (TRD-Net) based on message passing to estimate the joint policy divergence more accurately. The Multi-Agent Mirror descent policy algorithm with Trust region decomposition, called MAMT, is established by adjusting the trust-region of the local policies adaptively in an end-to-end manner. MAMT can approximately constrain the consecutive joint policies' divergence to satisfy $δ$-stationarity and alleviate the non-stationarity problem. Our method can bring noticeable and stable performance improvement compared with baselines in cooperative tasks of different complexity.

cs.LG↗

arXiv2022

Multi-Agent Path Finding with Prioritized Communication Learning

Multi-agent pathfinding (MAPF) has been widely used to solve large-scale real-world problems, e.g., automation warehouses. The learning-based, fully decentralized framework has been introduced to alleviate real-time problems and simultaneously pursue optimal planning policy. However, existing methods might generate significantly more vertex conflicts (or collisions), which lead to a low success rate or more makespan. In this paper, we propose a PrIoritized COmmunication learning method (PICO), which incorporates the \textit{implicit} planning priorities into the communication topology within the decentralized multi-agent reinforcement learning framework. Assembling with the classic coupled planners, the implicit priority learning module can be utilized to form the dynamic communication topology, which also builds an effective collision-avoiding mechanism. PICO performs significantly better in large-scale MAPF tasks in success rates and collision rates than state-of-the-art learning-based planners.

cs.RO↗

arXiv2021

VMAgent: Scheduling Simulator for Reinforcement Learning

A novel simulator called VMAgent is introduced to help RL researchers better explore new methods, especially for virtual machine scheduling. VMAgent is inspired by practical virtual machine (VM) scheduling tasks and provides an efficient simulation platform that can reflect the real situations of cloud computing. Three scenarios (fading, recovering, and expansion) are concluded from practical cloud computing and corresponds to many reinforcement learning challenges (high dimensional state and action spaces, high non-stationarity, and life-long demand). VMAgent provides flexible configurations for RL researchers to design their customized scheduling environments considering different problem features. From the VM scheduling perspective, VMAgent also helps to explore better learning-based scheduling solutions.

cs.LG↗

arXiv2021

Interactive Medical Image Segmentation with Self-Adaptive Confidence Calibration

Medical image segmentation is one of the fundamental problems for artificial intelligence-based clinical decision systems. Current automatic medical image segmentation methods are often failed to meet clinical requirements. As such, a series of interactive segmentation algorithms are proposed to utilize expert correction information. However, existing methods suffer from some segmentation refining failure problems after long-term interactions and some cost problems from expert annotation, which hinder clinical applications. This paper proposes an interactive segmentation framework, called interactive MEdical segmentation with self-adaptive Confidence CAlibration (MECCA), by introducing the corrective action evaluation, which combines the action-based confidence learning and multi-agent reinforcement learning (MARL). The evaluation is established through a novel action-based confidence network, and the corrective actions are obtained from MARL. Based on the confidential information, a self-adaptive reward function is designed to provide more detailed feedback, and a simulated label generation mechanism is proposed on unsupervised data to reduce over-reliance on labeled data. Experimental results on various medical image datasets have shown the significant performance of the proposed algorithm.

cs.CV↗

arXiv2021

HMRL: Hyper-Meta Learning for Sparse Reward Reinforcement Learning Problem

In spite of the success of existing meta reinforcement learning methods, they still have difficulty in learning a meta policy effectively for RL problems with sparse reward. In this respect, we develop a novel meta reinforcement learning framework called Hyper-Meta RL(HMRL), for sparse reward RL problems. It is consisted with three modules including the cross-environment meta state embedding module which constructs a common meta state space to adapt to different environments; the meta state based environment-specific meta reward shaping which effectively extends the original sparse reward trajectory by cross-environmental knowledge complementarity and as a consequence the meta policy achieves better generalization and efficiency with the shaped meta reward. Experiments with sparse-reward environments show the superiority of HMRL on both transferability and policy learning efficiency.

cs.LG↗

arXiv2021

Structured Diversification Emergence via Reinforced Organization Control and Hierarchical Consensus Learning

When solving a complex task, humans will spontaneously form teams and to complete different parts of the whole task, respectively. Meanwhile, the cooperation between teammates will improve efficiency. However, for current cooperative MARL methods, the cooperation team is constructed through either heuristics or end-to-end blackbox optimization. In order to improve the efficiency of cooperation and exploration, we propose a structured diversification emergence MARL framework named {\sc{Rochico}} based on reinforced organization control and hierarchical consensus learning. {\sc{Rochico}} first learns an adaptive grouping policy through the organization control module, which is established by independent multi-agent reinforcement learning. Further, the hierarchical consensus module based on the hierarchical intentions with consensus constraint is introduced after team formation. Simultaneously, utilizing the hierarchical consensus module and a self-supervised intrinsic reward enhanced decision module, the proposed cooperative MARL algorithm {\sc{Rochico}} can output the final diversified multi-agent cooperative policy. All three modules are organically combined to promote the structured diversification emergence. Comparative experiments on four large-scale cooperation tasks show that {\sc{Rochico}} is significantly better than the current SOTA algorithms in terms of exploration efficiency and cooperation strength.

cs.LG↗

arXiv2020

Electrical detection of the inverse Edelstein effect on the surface of SmB$_6$

We report the measurement of spin current induced charge accumulation, the inverse Edelstein effect (IEE), on the surface of a candidate topological Kondo insulator SmB6 single crystal. Robust surface conduction channel of SmB6 has been shown to exhibit large degree of spin-momentum locking, and spin polarized current through an external ferromagnetic contact induces the spin dependent charge accumulation on the surface of SmB6. The dependences of the IEE signal on the bias current, an external magnetic field direction and temperature are consistent with the anticlockwise spin texture for the surface band in SmB6 in the momentum space, and the direction and magnitude of the effect compared with the normal Edelstein signal are clearly explained by the Onsager reciprocal relation. Furthermore, we estimate spin-to-charge conversion efficiency, the IEE length, as 4.46 nm that is an order of magnitude larger than the efficiency found in other typical Rashba interfaces, implying that the Rashba contribution to the IEE signal could be small. Building upon existing reports on the surface charge and spin conduction nature on this material, our results provide additional evidence that the surface of SmB6 supports spin polarized conduction channel.

physics.app-ph↗

arXiv2020

FDA3 : Federated Defense Against Adversarial Attacks for Cloud-Based IIoT Applications

Along with the proliferation of Artificial Intelligence (AI) and Internet of Things (IoT) techniques, various kinds of adversarial attacks are increasingly emerging to fool Deep Neural Networks (DNNs) used by Industrial IoT (IIoT) applications. Due to biased training data or vulnerable underlying models, imperceptible modifications on inputs made by adversarial attacks may result in devastating consequences. Although existing methods are promising in defending such malicious attacks, most of them can only deal with limited existing attack types, which makes the deployment of large-scale IIoT devices a great challenge. To address this problem, we present an effective federated defense approach named FDA3 that can aggregate defense knowledge against adversarial examples from different sources. Inspired by federated learning, our proposed cloud-based architecture enables the sharing of defense capabilities against different attacks among IIoT devices. Comprehensive experimental results show that the generated DNNs by our approach can not only resist more malicious attacks than existing attack-specific adversarial training methods, but also can prevent IIoT applications from new attacks.

cs.LG↗

arXiv2020

Learning Structured Communication for Multi-agent Reinforcement Learning

This work explores the large-scale multi-agent communication mechanism under a multi-agent reinforcement learning (MARL) setting. We summarize the general categories of topology for communication structures in MARL literature, which are often manually specified. Then we propose a novel framework termed as Learning Structured Communication (LSC) by using a more flexible and efficient communication topology. Our framework allows for adaptive agent grouping to form different hierarchical formations over episodes, which is generated by an auxiliary task combined with a hierarchical routing protocol. Given each formed topology, a hierarchical graph neural network is learned to enable effective message information generation and propagation among inter- and intra-group communications. In contrast to existing communication mechanisms, our method has an explicit while learnable design for hierarchical communication. Experiments on challenging tasks show the proposed LSC enjoys high communication efficiency, scalability, and global cooperation capability.

cs.LG↗

arXiv2019

Iteratively-Refined Interactive 3D Medical Image Segmentation with Multi-Agent Reinforcement Learning

Existing automatic 3D image segmentation methods usually fail to meet the clinic use. Many studies have explored an interactive strategy to improve the image segmentation performance by iteratively incorporating user hints. However, the dynamic process for successive interactions is largely ignored. We here propose to model the dynamic process of iterative interactive image segmentation as a Markov decision process (MDP) and solve it with reinforcement learning (RL). Unfortunately, it is intractable to use single-agent RL for voxel-wise prediction due to the large exploration space. To reduce the exploration space to a tractable size, we treat each voxel as an agent with a shared voxel-level behavior strategy so that it can be solved with multi-agent reinforcement learning. An additional advantage of this multi-agent model is to capture the dependency among voxels for segmentation task. Meanwhile, to enrich the information of previous segmentations, we reserve the prediction uncertainty in the state space of MDP and derive an adjustment action space leading to a more precise and finer segmentation. In addition, to improve the efficiency of exploration, we design a relative cross-entropy gain-based reward to update the policy in a constrained direction. Experimental results on various medical datasets have shown that our method significantly outperforms existing state-of-the-art methods, with the advantage of fewer interactions and a faster convergence.

cs.CV↗

arXiv2019

Heterogeneous Graph-based Knowledge Transfer for Generalized Zero-shot Learning

Generalized zero-shot learning (GZSL) tackles the problem of learning to classify instances involving both seen classes and unseen ones. The key issue is how to effectively transfer the model learned from seen classes to unseen classes. Existing works in GZSL usually assume that some prior information about unseen classes are available. However, such an assumption is unrealistic when new unseen classes appear dynamically. To this end, we propose a novel heterogeneous graph-based knowledge transfer method (HGKT) for GZSL, agnostic to unseen classes and instances, by leveraging graph neural network. Specifically, a structured heterogeneous graph is constructed with high-level representative nodes for seen classes, which are chosen through Wasserstein barycenter in order to simultaneously capture inter-class and intra-class relationship. The aggregation and embedding functions can be learned through graph neural network, which can be used to compute the embeddings of unseen classes by transferring the knowledge from their neighbors. Extensive experiments on public benchmark datasets show that our method achieves state-of-the-art results.

cs.LG↗

arXiv2019

Definitive Surface Magnetotransport Study of SmB$_{6}$

After the theoretical prediction that SmB$_6$ is a topological Kondo insulator, there has been an explosion of studies on the SmB$_6$ surface. However, there is not yet an agreement on even the most basic quantities such as the surface carrier density and mobility. In this paper, we carefully revisit Corbino disk magnetotransport studies to find those surface transport parameters. We first show that subsurface cracks exist in the SmB$_6$ crystals, arising both from surface preparation and during the crystal growth. We provide evidence that these hidden subsurface cracks are additional conduction channels, and the large disagreement between earlier surface SmB$_6$ studies may originate from previous interpretations not taking this extra conduction path into account. We provide an update of a more reliable magnetotransport data than the previous one (Phys. Rev. B 92, 115110) and find that the orders-of-magnitude large disagreements in carrier density and mobility come from the surface preparation and the transport geometry rather than the intrinsic sample quality. From this magnetotransport study, we find an updated estimate of the carrier density and mobility of 2.71$\times$10$^{13}$ (1/cm$^2$) and 104.5 (cm$^{2}$/V$\cdot$sec), respectively. We compare our results with other studies of the SmB$_6$ surface. By this comparison, we provide insight into the disagreements and agreements of the previously reported angle-resolved photoemission spectroscopy, scanning tunneling microscopy, and magnetotorque quantum oscillations measurements.

cond-mat.mes-hall↗

arXiv2019

A Fast Proximal Point Method for Computing Exact Wasserstein Distance

Wasserstein distance plays increasingly important roles in machine learning, stochastic programming and image processing. Major efforts have been under way to address its high computational complexity, some leading to approximate or regularized variations such as Sinkhorn distance. However, as we will demonstrate, regularized variations with large regularization parameter will degradate the performance in several important machine learning applications, and small regularization parameter will fail due to numerical stability issues with existing algorithms. We address this challenge by developing an Inexact Proximal point method for exact Optimal Transport problem (IPOT) with the proximal operator approximately evaluated at each iteration using projections to the probability simplex. The algorithm (a) converges to exact Wasserstein distance with theoretical guarantee and robust regularization parameter selection, (b) alleviates numerical stability issue, (c) has similar computational complexity to Sinkhorn, and (d) avoids the shrinking problem when apply to generative models. Furthermore, a new algorithm is proposed based on IPOT to obtain sharper Wasserstein barycenter.

stat.ML↗

arXiv2019

Observation of Two Collapsed Phases in CaRbFe4As4

We report the observation of the pressure-induced, fully-collapsed tetragonal phase of CaRbFe4As4 for P~ 22 GPa via high-pressure x-ray diffraction and magnetotransport measurements. The x-ray measurements, along with resistivity measurements, show that there is an initial half-collapsed tetragonal phase for 6 < P < 22 GPa, in which superconductivity is continuously suppressed from Tc= 35K at P= 3.1 GPa to Tc <2K for P >17.2 GPa, as well as signs of the fully-collapsed tetragonal phase near P=22 GPa. Density functional calculations suggest that both of these transitions are driven by increased As-As bonding, first across the Ca layer, and then at the second transition, across the Rb layer. Although electrical resistivity measurements in the fully-collapsed tetragonal phase do not show superconductivity, there is a change in the slope of both the Hall coefficient and the longitudinal resistance near 22 GPa, suggesting a strong correlation between the electronic and lattice degrees of freedom in this new iron-based superconductor.

cond-mat.str-el↗

arXiv2018

Imaging emergent heavy Dirac fermions of a topological Kondo insulator

Kondo insulators are primary candidates in the search for strongly correlated topological quantum phases, which may host topological order, fractionalization, and non-Abelian statistics. Within some Kondo insulators, the hybridization gap is predicted to protect a nontrivial topological invariant and to harbor emergent heavy Dirac fermion surface modes. We use high-energy-resolution spectroscopic imaging in real and momentum space on the Kondo insulator, SmB$_6$. On cooling through $T^*_Δ\approx$ 35 K we observe the opening of an insulating gap that expands to $Δ\approx$ 10 meV at 2 K. Within the gap, we image the formation of linearly dispersing surface states with effective masses reaching $m^* = (410\pm20)m_e$. We thus demonstrate existence of a strongly correlated topological Kondo insulator phase hosting the heaviest known Dirac fermions.

cond-mat.str-el↗

arXiv2018

Perturbation techniques for convergence analysis of proximal gradient method and other first-order algorithms via variational analysis

We develop new perturbation techniques for conducting convergence analysis of various first-order algorithms for a class of nonsmooth optimization problems. We consider the iteration scheme of an algorithm to construct a perturbed stationary point set-valued map, and define the perturbing parameter by the difference of two consecutive iterates. Then, we show that the calmness condition of the induced set-valued map, together with a local version of the proper separation of stationary value condition, is a sufficient condition to ensure the linear convergence of the algorithm. The equivalence of the calmness condition to the one for the canonically perturbed stationary point set-valued map is proved, and this equivalence allows us to derive some sufficient conditions for calmness by using some recent developments in variational analysis. These sufficient conditions are different from existing results (especially, those error-bound-based ones) in that they can be easily verified for many concrete application models. Our analysis is focused on the fundamental proximal gradient (PG) method, and it enables us to show that any accumulation of the sequence generated by the PG method must be a stationary point in terms of the proximal subdifferential, instead of the limiting subdifferential. This result finds the surprising fact that the solution quality found by the PG method is in general superior. Our analysis also leads to some improvement for the linear convergence results of the PG method in the convex case. The new perturbation technique can be conveniently used to derive linear rate convergence of a number of other first-order methods including the well-known alternating direction method of multipliers and primal-dual hybrid gradient method, under mild assumptions.

math.OC↗

arXiv2018

Electrical detection of surface spin polarization of candidate topological Kondo insulator SmB6

The Kondo insulator compound SmB6 has emerged as a strong candidate for the realization of a topologically nontrivial state in a strongly correlated system, a topological Kondo insulator, which can be a novel platform for investigating the interplay between nontrivial topology and emergent correlation driven phenomena in solid state systems. Electronic transport measurements on this material, however, so far showed only the robust surface dominated charge conduction at low temperatures, lacking evidence of its connection to the topological nature by showing, for example, spin polarization due to spin momentum locking. Here, we find evidence for surface state spin polarization by electrical detection of a current induced spin chemical potential difference on the surface of a SmB6 single crystal. We clearly observe a surface dominated spin voltage, which is proportional to the projection of the spin polarization onto the contact magnetization, is determined by the direction and magnitude of the charge current and is strongly temperature dependent due to the crossover from surface to bulk conduction. We estimate the lower bound of the surface state net spin polarization as 15 percent based on the quantum transport model providing direct evidence that SmB6 supports metallic spin helical surface states.

cond-mat.str-el↗