arXiv subjects

Hao Zhou

Publications and source records attributed to Hao Zhou.

At least 361 records · Page 20Linked to original sources

arXiv2023

Hierarchical Reinforcement Learning Based Traffic Steering in Multi-RAT 5G Deployments

In 5G non-standalone mode, an intelligent traffic steering mechanism can vastly aid in ensuring smooth user experience by selecting the best radio access technology (RAT) from a multi-RAT environment for a specific traffic flow. In this paper, we propose a novel load-aware traffic steering algorithm based on hierarchical reinforcement learning (HRL) while satisfying diverse QoS requirements of different traffic types. HRL can significantly increase system performance using a bi-level architecture having a meta-controller and a controller. In our proposed method, the meta-controller provides an appropriate threshold for load balancing, while the controller performs traffic admission to an appropriate RAT in the lower level. Simulation results show that HRL outperforms a Deep Q-Learning (DQN) and a threshold-based heuristic baseline with 8.49%, 12.52% higher average system throughput and 27.74%, 39.13% lower network delay, respectively.

cs.NI↗

arXiv2023

Traffic Steering for 5G Multi-RAT Deployments using Deep Reinforcement Learning

In 5G non-standalone mode, traffic steering is a critical technique to take full advantage of 5G new radio while optimizing dual connectivity of 5G and LTE networks in multiple radio access technology (RAT). An intelligent traffic steering mechanism can play an important role to maintain seamless user experience by choosing appropriate RAT (5G or LTE) dynamically for a specific user traffic flow with certain QoS requirements. In this paper, we propose a novel traffic steering mechanism based on Deep Q-learning that can automate traffic steering decisions in a dynamic environment having multiple RATs, and maintain diverse QoS requirements for different traffic classes. The proposed method is compared with two baseline algorithms: a heuristic-based algorithm and Q-learningbased traffic steering. Compared to the Q-learning and heuristic baselines, our results show that the proposed algorithm achieves better performance in terms of 6% and 10% higher average system throughput, and 23% and 33% lower network delay, respectively.

cs.NI↗

arXiv2023

Hierarchical Reinforcement Learning for RIS-Assisted Energy-Efficient RAN

Reconfigurable intelligent surface (RIS) is emerging as a promising technology to boost the energy efficiency (EE) of 5G beyond and 6G networks. Inspired by this potential, in this paper, we investigate the RIS-assisted energy-efficient radio access networks (RAN). In particular, we combine RIS with sleep control techniques, and develop a hierarchical reinforcement learning (HRL) algorithm for network management. In HRL, the meta-controller decides the on/off status of the small base stations (SBSs) in heterogeneous networks, while the sub-controller can change the transmission power levels of SBSs to save energy. The simulations show that the RIS-assisted sleep control can achieve significantly lower energy consumption, higher throughput, and more than doubled energy efficiency than no-RIS conditions.

eess.SP↗

arXiv2023

Seismic Wave Scattering and Dissipation in Fractured Shales

Seismic attenuation in granular porous media is of paramount importance in rock physics and seismology. Unlike sandstones, shales are mixtures of sand grains and clays with extremely low porosity and permeability. Swelling of clays upon wetting induce micro-cracks at grain-clay interfaces and results in the strong elastic wave scattering. Such scattering prevents adequate measurements of the absorption from ballistic wave attenuations. Here we infer this intrinsic attenuation from multiply scattered waves as in seismology and ultrasonics. We find that increasing confining pressure reduces the scattering attenuation by micro-crack closure but increases surprisingly the absorption, likely due to the viscous dissipation involved with more liquids adsorbed in clays and at grain surfaces. Also, we observe that cyclic heating and cooling causes the shrinkage of clays and the growth of microcracks as well as the nucleation of macro-fractures. This leads to a predominant chaotic reverberation in this fractured shale. Numerical simulations based on X-ray tomography of the fractured sample confirm the multiple scattering behavior and reveal the increase of a characteristic length from an initial intact to a finally fractured shale. This study helps to improve acoustic techniques for multiscale exploration of gas and oil in shales and other fractured rocks.

cond-mat.dis-nn↗

arXiv2022

Knowledge Transfer and Reuse: A Case Study of AI-enabled Resource Management in RAN Slicing

An efficient resource management scheme is critical to enable network slicing in 5G networks and in envisioned 6G networks, and artificial intelligence (AI) techniques offer promising solutions. Considering the rapidly emerging new machine learning techniques, such as graph learning, federated learning, and transfer learning, a timely survey is needed to provide an overview of resource management and network slicing techniques of AI-enabled wireless networks. This article provides such a survey along with an application of knowledge transfer in radio access network (RAN) slicing. In particular, we firs provide some background on resource management and network slicing, and review relevant state-of-the-art AI and machine learning (ML) techniques and their applications. Then, we introduce our AI-enabled knowledge transfer and reuse-based resource management (AKRM) scheme, where we apply transfer learning to improve system performance. Compared with most existing works, which focus on the training of standalone agents from scratch, the main difference of AKRM lies in its knowledge transfer and reuse capability between different tasks. Our paper aims to be a roadmap for researchers to use knowledge transfer schemes in AI-enabled wireless networks, and we provide a case study over the resource allocation problem in RAN slicing.

cs.NI↗

arXiv2022

GRB 080503: A very early blue kilonova and an adjacent non-thermal radiation component

The temporal behavior of the very dim optical afterglow of GRB 080503 is at odds with the regular forward shock afterglow model and a sole kilonova component responsible for optical emission has been speculated in some literature. Here we analyze the optical afterglow data available in archive and construct time-resolved spectra. The significant detection by Keck-I in {\it G/R} bands at $t\sim 3$ day, which has not been reported before, as well as the simultaneous Gemini-North {\it r} band measurement, are in favor of a power-law spectrum that is well consistent with the optical to X-ray spectrum measured at $t\sim 4.5$ day. However, for $t\leq 2$ day, the spectra are thermal-like and a straightforward interpretation is a kilonova emission from a neutron star merger, making it, possibly, the first detection of a very early kilonova signal at $t\sim 0.05$ day. A non-thermal nature of optical emission at late times ($t\sim 2$ day), anyhow, can not be ruled out because of the large uncertainty of the {\it g}-band data. We also propose to classify the neutron star merger induced optical transients, according to the temporal behaviors of the kilonova and the non-thermal afterglow emission, into four types. GRB 080503 would then represent the first observation of a sub-group of neutron star merger driven optical transients (i.e., Type IV) consisting of an early blue kilonova and an adjacent non-thermal afterglow radiation.

astro-ph.HE↗

arXiv2022

CTRLEval: An Unsupervised Reference-Free Metric for Evaluating Controlled Text Generation

Existing reference-free metrics have obvious limitations for evaluating controlled text generation models. Unsupervised metrics can only provide a task-agnostic evaluation result which correlates weakly with human judgments, whereas supervised ones may overfit task-specific data with poor generalization ability to other datasets. In this paper, we propose an unsupervised reference-free metric called CTRLEval, which evaluates controlled text generation from different aspects by formulating each aspect into multiple text infilling tasks. On top of these tasks, the metric assembles the generation probabilities from a pre-trained language model without any model training. Experimental results show that our metric has higher correlations with human judgments than other baselines, while obtaining better generalization of evaluating generated texts from different models and with different qualities.

cs.CL↗

arXiv2022

Direct Heterogeneous Causal Learning for Resource Allocation Problems in Marketing

Marketing is an important mechanism to increase user engagement and improve platform revenue, and heterogeneous causal learning can help develop more effective strategies. Most decision-making problems in marketing can be formulated as resource allocation problems and have been studied for decades. Existing works usually divide the solution procedure into two fully decoupled stages, i.e., machine learning (ML) and operation research (OR) -- the first stage predicts the model parameters and they are fed to the optimization in the second stage. However, the error of the predicted parameters in ML cannot be respected and a series of complex mathematical operations in OR lead to the increased accumulative errors. Essentially, the improved precision on the prediction parameters may not have a positive correlation on the final solution due to the side-effect from the decoupled design. In this paper, we propose a novel approach for solving resource allocation problems to mitigate the side-effects. Our key intuition is that we introduce the decision factor to establish a bridge between ML and OR such that the solution can be directly obtained in OR by only performing the sorting or comparison operations on the decision factor. Furthermore, we design a customized loss function that can conduct direct heterogeneous causal learning on the decision factor, an unbiased estimation of which can be guaranteed when the loss converges. As a case study, we apply our approach to two crucial problems in marketing: the binary treatment assignment problem and the budget allocation problem with multiple treatments. Both large-scale simulations and online A/B Tests demonstrate that our approach achieves significant improvement compared with state-of-the-art.

cs.LG↗

arXiv2022

Topological bulk solitons in a nonlinear photonic Chern insulator

Nonlinearities in lattices with topological band structures can induce topological interfaces in the bulk of structures and give rise to bulk solitons in the topological bandgaps. Here we study a photonic Chern insulator with saturable nonlinearity and show the existence of topological bulk solitons. The fundamental bulk solitons exhibit as semi-vortex solitons, where only one pseudospin component has a nonzero vorticity. The bulk solitons have equal angular momentum at different valleys. This phenomenon is a direct outcome of the topology of the linear host lattice and the angular momentum can be changed by switching the sign of the nonlinearity. The bulk solitons bifurcate from the linear bulk band edge and terminate when their powers saturate. We find that these bulk solitons are stable within the whole spectrum range. Moreover, these bulk solitons are robust against lattice disorders both from on-site energies and hopping amplitudes. Our work extends the study of Chern insulators into the nonlinear regime and highlights the interplay between topology and nonlinearity.

physics.optics↗

arXiv2022

ROSE: Robust Selective Fine-tuning for Pre-trained Language Models

Even though the large-scale language models have achieved excellent performances, they suffer from various adversarial attacks. A large body of defense methods has been proposed. However, they are still limited due to redundant attack search spaces and the inability to defend against various types of attacks. In this work, we present a novel fine-tuning approach called \textbf{RO}bust \textbf{SE}letive fine-tuning (\textbf{ROSE}) to address this issue. ROSE conducts selective updates when adapting pre-trained models to downstream tasks, filtering out invaluable and unrobust updates of parameters. Specifically, we propose two strategies: the first-order and second-order ROSE for selecting target robust parameters. The experimental results show that ROSE achieves significant improvements in adversarial robustness on various downstream NLP tasks, and the ensemble method even surpasses both variants above. Furthermore, ROSE can be easily incorporated into existing fine-tuning methods to improve their adversarial robustness further. The empirical analysis confirms that ROSE eliminates unrobust spurious updates during fine-tuning, leading to solutions corresponding to flatter and wider optima than the conventional method. Code is available at \url{https://github.com/jiangllan/ROSE}.

cs.CL↗

arXiv2022

Prompt-based Connective Prediction Method for Fine-grained Implicit Discourse Relation Recognition

Due to the absence of connectives, implicit discourse relation recognition (IDRR) is still a challenging and crucial task in discourse analysis. Most of the current work adopted multi-task learning to aid IDRR through explicit discourse relation recognition (EDRR) or utilized dependencies between discourse relation labels to constrain model predictions. But these methods still performed poorly on fine-grained IDRR and even utterly misidentified on most of the few-shot discourse relation classes. To address these problems, we propose a novel Prompt-based Connective Prediction (PCP) method for IDRR. Our method instructs large-scale pre-trained models to use knowledge relevant to discourse relation and utilizes the strong correlation between connectives and discourse relation to help the model recognize implicit discourse relations. Experimental results show that our method surpasses the current state-of-the-art model and achieves significant improvements on those fine-grained few-shot discourse relation. Moreover, our approach is able to be transferred to EDRR and obtain acceptable results. Our code is released in https://github.com/zh-i9/PCP-for-IDRR.

cs.CL↗

arXiv2022

PARAGEN : A Parallel Generation Toolkit

PARAGEN is a PyTorch-based NLP toolkit for further development on parallel generation. PARAGEN provides thirteen types of customizable plugins, helping users to experiment quickly with novel ideas across model architectures, optimization, and learning strategies. We implement various features, such as unlimited data loading and automatic model selection, to enhance its industrial usage. ParaGen is now deployed to support various research and industry applications at ByteDance. PARAGEN is available at https://github.com/bytedance/ParaGen.

cs.CL↗

arXiv2022

Zero-Shot 3D Drug Design by Sketching and Generating

Drug design is a crucial step in the drug discovery cycle. Recently, various deep learning-based methods design drugs by generating novel molecules from scratch, avoiding traversing large-scale drug libraries. However, they depend on scarce experimental data or time-consuming docking simulation, leading to overfitting issues with limited training data and slow generation speed. In this study, we propose the zero-shot drug design method DESERT (Drug dEsign by SkEtching and geneRaTing). Specifically, DESERT splits the design process into two stages: sketching and generating, and bridges them with the molecular shape. The two-stage fashion enables our method to utilize the large-scale molecular database to reduce the need for experimental data and docking simulation. Experiments show that DESERT achieves a new state-of-the-art at a fast speed.

cs.CE↗

arXiv2022

IMB-NAS: Neural Architecture Search for Imbalanced Datasets

Class imbalance is a ubiquitous phenomenon occurring in real world data distributions. To overcome its detrimental effect on training accurate classifiers, existing work follows three major directions: class re-balancing, information transfer, and representation learning. In this paper, we propose a new and complementary direction for improving performance on long tailed datasets - optimizing the backbone architecture through neural architecture search (NAS). We find that an architecture's accuracy obtained on a balanced dataset is not indicative of good performance on imbalanced ones. This poses the need for a full NAS run on long tailed datasets which can quickly become prohibitively compute intensive. To alleviate this compute burden, we aim to efficiently adapt a NAS super-network from a balanced source dataset to an imbalanced target one. Among several adaptation strategies, we find that the most effective one is to retrain the linear classification head with reweighted loss, while freezing the backbone NAS super-network trained on a balanced source dataset. We perform extensive experiments on multiple datasets and provide concrete insights to optimize architectures for long tailed datasets.

cs.LG↗

arXiv2022

Towards Regression-Free Neural Networks for Diverse Compute Platforms

With the shift towards on-device deep learning, ensuring a consistent behavior of an AI service across diverse compute platforms becomes tremendously important. Our work tackles the emergent problem of reducing predictive inconsistencies arising as negative flips: test samples that are correctly predicted by a less accurate model, but incorrectly by a more accurate one. We introduce REGression constrained Neural Architecture Search (REG-NAS) to design a family of highly accurate models that engender fewer negative flips. REG-NAS consists of two components: (1) A novel architecture constraint that enables a larger model to contain all the weights of the smaller one thus maximizing weight sharing. This idea stems from our observation that larger weight sharing among networks leads to similar sample-wise predictions and results in fewer negative flips; (2) A novel search reward that incorporates both Top-1 accuracy and negative flips in the architecture search metric. We demonstrate that \regnas can successfully find desirable architectures with few negative flips in three popular architecture search spaces. Compared to the existing state-of-the-art approach, REG-NAS enables 33-48% relative reduction of negative flips.

cs.CV↗

arXiv2022

Congested Urban Networks Tend to Be Insensitive to Signal Settings: Implications for Learning-Based Control

This paper highlights several properties of large urban networks that can have an impact on machine learning methods applied to traffic signal control. In particular, we show that the average network flow tends to be independent of the signal control policy as density increases. This property, which so far has remained under the radar, implies that deep reinforcement learning (DRL) methods becomes ineffective when trained under congested conditions, and might explain DRL's limited success for traffic signal control. Our results apply to all possible grid networks thanks to a parametrization based on two network parameters: the ratio of the expected distance between consecutive traffic lights to the expected green time, and the turning probability at intersections. Networks with different parameters exhibit very different responses to traffic signal control. Notably, we found that no control (i.e. random policy) can be an effective control strategy for a surprisingly large family of networks. The impact of the turning probability turned out to be very significant both for baseline and for DRL policies. It also explains the loss of symmetry observed for these policies, which is not captured by existing theories that rely on corridor approximations without turns. Our findings also suggest that supervised learning methods have enormous potential as they require very little examples to produce excellent policies.

cs.LG↗

arXiv2022

Learning from Peers: Deep Transfer Reinforcement Learning for Joint Radio and Cache Resource Allocation in 5G RAN Slicing

Network slicing is a critical technique for 5G communications that covers radio access network (RAN), edge, transport and core slicing.The evolving network architecture requires the orchestration of multiple network resources such as radio and cache resources. In recent years, machine learning (ML) techniques have been widely applied for network management. However, most existing works do not take advantage of the knowledge transfer capability in ML. In this paper, we propose a deep transfer reinforcement learning (DTRL) scheme for joint radio and cache resource allocation to serve 5G RAN slicing. We first define a hierarchical architecture for joint resource allocation. Then we propose two DTRL algorithms: Q-value-based deep transfer reinforcement learning (QDTRL) and action selection-based deep transfer reinforcement learning (ADTRL). In the proposed schemes, learner agents utilize expert agents' knowledge to improve their performance on current tasks. The proposed algorithms are compared with both the model-free exploration bonus deep Q-learning (EB-DQN) and the model-based priority proportional fairness and time-to-live (PPF-TTL) algorithms. Compared with EB-DQN, our proposed DTRL-based method presents 21.4% lower delay for Ultra Reliable Low Latency Communications (URLLC) slice and 22.4% higher throughput for enhanced Mobile Broad Band (eMBB) slice, while achieving significantly faster convergence than EB-DQN. Moreover, 40.8% lower URLLC delay and 59.8% higher eMBB throughput are observed with respect to PPF-TTL.

eess.SY↗

arXiv2022

Manual-Guided Dialogue for Flexible Conversational Agents

How to build and use dialogue data efficiently, and how to deploy models in different domains at scale can be two critical issues in building a task-oriented dialogue system. In this paper, we propose a novel manual-guided dialogue scheme to alleviate these problems, where the agent learns the tasks from both dialogue and manuals. The manual is an unstructured textual document that guides the agent in interacting with users and the database during the conversation. Our proposed scheme reduces the dependence of dialogue models on fine-grained domain ontology, and makes them more flexible to adapt to various domains. We then contribute a fully-annotated multi-domain dataset MagDial to support our scheme. It introduces three dialogue modeling subtasks: instruction matching, argument filling, and response generation. Modeling these subtasks is consistent with the human agent's behavior patterns. Experiments demonstrate that the manual-guided dialogue scheme improves data efficiency and domain scalability in building dialogue systems. The dataset and benchmark will be publicly available for promoting future research.

cs.CL↗