arXiv subjects

Yuan Zhou

Publications and source records attributed to Yuan Zhou.

At least 163 records · Page 9Linked to original sources

arXiv2023

Network Revenue Management with Demand Learning and Fair Resource-Consumption Balancing

In addition to maximizing the total revenue, decision-makers in lots of industries would like to guarantee balanced consumption across different resources. For instance, in the retailing industry, ensuring a balanced consumption of resources from different suppliers enhances fairness and helps main a healthy channel relationship; in the cloud computing industry, resource-consumption balance helps increase customer satisfaction and reduce operational costs. Motivated by these practical needs, this paper studies the price-based network revenue management (NRM) problem with both demand learning and fair resource-consumption balancing. We introduce the regularized revenue, i.e., the total revenue with a balancing regularization, as our objective to incorporate fair resource-consumption balancing into the revenue maximization goal. We propose a primal-dual-type online policy with the Upper-Confidence-Bound (UCB) demand learning method to maximize the regularized revenue. We adopt several innovative techniques to make our algorithm a unified and computationally efficient framework for the continuous price set and a wide class of balancing regularizers. Our algorithm achieves a worst-case regret of $\widetilde O(N^{5/2}\sqrt{T})$, where $N$ denotes the number of products and $T$ denotes the number of time periods. Numerical experiments in a few NRM examples demonstrate the effectiveness of our algorithm in simultaneously achieving revenue maximization and fair resource-consumption balancing

stat.ML↗

arXiv2023

Explicit lump and line rogue wave solutions to a modified Hietarinta equation

Lump solutions are spatially rationally localized solutions which usually arise as solutions to higher dimensional nonlinear partial differential equations often possessing Hirota bilinear forms. Under some parameter constraint, these solutions may lead to rogue wave solutions. In this article, we study lump and rogue wave solutions of a new nonlinear non-evolutionary equation in 2+1 dimensions with the aid of a computer algebra system. We present illustrative examples and analyze the dynamical behavior of the solutions using graphical representations

nlin.SI↗

arXiv2023

Fairness-aware Online Price Discrimination with Nonparametric Demand Models

Price discrimination, which refers to the strategy of setting different prices for different customer groups, has been widely used in online retailing. Although it helps boost the collected revenue for online retailers, it might create serious concerns about fairness, which even violates the regulation and laws. This paper studies the problem of dynamic discriminatory pricing under fairness constraints. In particular, we consider a finite selling horizon of length $T$ for a single product with two groups of customers. Each group of customers has its unknown demand function that needs to be learned. For each selling period, the seller determines the price for each group and observes their purchase behavior. While existing literature mainly focuses on maximizing revenue, ensuring fairness among different customers has not been fully explored in the dynamic pricing literature. This work adopts the fairness notion from Cohen et al. (2022). For price fairness, we propose an optimal dynamic pricing policy regarding regret, which enforces the strict price fairness constraint. In contrast to the standard $\sqrt{T}$-type regret in online learning, we show that the optimal regret in our case is $\tilde{O}(T^{4/5})$. We further extend our algorithm to a more general notion of fairness, which includes demand fairness as a special case. To handle this general class, we propose a soft fairness constraint and develop a dynamic pricing policy that achieves $\tilde{O}(T^{4/5})$ regret. We also demonstrate that our algorithmic techniques can be adapted to more general scenarios such as fairness among multiple groups of customers.

cs.LG↗

arXiv2023

BehAVExplor: Behavior Diversity Guided Testing for Autonomous Driving Systems

Testing Autonomous Driving Systems (ADSs) is a critical task for ensuring the reliability and safety of autonomous vehicles. Existing methods mainly focus on searching for safety violations while the diversity of the generated test cases is ignored, which may generate many redundant test cases and failures. Such redundant failures can reduce testing performance and increase failure analysis costs. In this paper, we present a novel behavior-guided fuzzing technique (BehAVExplor) to explore the different behaviors of the ego vehicle (i.e., the vehicle controlled by the ADS under test) and detect diverse violations. Specifically, we design an efficient unsupervised model, called BehaviorMiner, to characterize the behavior of the ego vehicle. BehaviorMiner extracts the temporal features from the given scenarios and performs a clustering-based abstraction to group behaviors with similar features into abstract states. A new test case will be added to the seed corpus if it triggers new behaviors (e.g., cover new abstract states). Due to the potential conflict between the behavior diversity and the general violation feedback, we further propose an energy mechanism to guide the seed selection and the mutation. The energy of a seed quantifies how good it is. We evaluated BehAVExplor on Apollo, an industrial-level ADS, and LGSVL simulation environment. Empirical evaluation results show that BehAVExplor can effectively find more diverse violations than the state-of-the-art.

cs.SE↗

arXiv2023

Threshold-independent method for single-shot readout of spin qubits in semiconductor quantum dots

The single-shot readout data process is essential for the realization of high-fidelity qubits and fault-tolerant quantum algorithms in semiconductor quantum dots. However, the fidelity and visibility of the readout process is sensitive to the choice of the thresholds and limited by the experimental hardware. By demonstrating the linear dependence between the measured spin state probabilities and readout visibilities along with dark counts, we describe an alternative threshold-independent method for the single-shot readout of spin qubits in semiconductor quantum dots. We can obtain the extrapolated spin state probabilities of the prepared probabilities of the excited spin state through the threshold-independent method. Then, we analyze the corresponding errors of the method, finding that errors of the extrapolated probabilities cannot be neglected with no constraints on the readout time and threshold voltage. Therefore, by limiting the readout time and threshold voltage we ensure the accuracy of the extrapolated probability. Then, we prove that the efficiency and robustness of this method is 60 times larger than that of the most commonly used method. Moreover, we discuss the influence of the electron temperature on the effective area with a fixed external magnetic field and provide a preliminary demonstration for a single-shot readout up to 0.7 K/1.5T in the future.

cond-mat.mes-hall↗

arXiv2023

On-chip arrayed waveguide grating fabricated on thin film lithium niobate

We design an on-chip 8-channel TFLN AWG and fabricate the device using photolithography assisted chemo-mechanical etching (PLACE) technique. We experimentally measure the transmission of the fabricated TFLN AWG near the central wavelength of 1550 nm. We obtain an on-chip loss as low as 3.32 dB, a single-channel bandwidth of 1.6 nm and a total-channel bandwidth of 12.8 nm. The crosstalk between adjacent channels was measured to be below -7.01 dB within the wavelength range from 1543 nm to 1558 nm, and the crosstalk between non-adjacent channels was below -15 dB.

physics.optics↗

arXiv2023

ABC-KD: Attention-Based-Compression Knowledge Distillation for Deep Learning-Based Noise Suppression

Noise suppression (NS) models have been widely applied to enhance speech quality. Recently, Deep Learning-Based NS, which we denote as Deep Noise Suppression (DNS), became the mainstream NS method due to its excelling performance over traditional ones. However, DNS models face 2 major challenges for supporting the real-world applications. First, high-performing DNS models are usually large in size, causing deployment difficulties. Second, DNS models require extensive training data, including noisy audios as inputs and clean audios as labels. It is often difficult to obtain clean labels for training DNS models. We propose the use of knowledge distillation (KD) to resolve both challenges. Our study serves 2 main purposes. To begin with, we are among the first to comprehensively investigate mainstream KD techniques on DNS models to resolve the two challenges. Furthermore, we propose a novel Attention-Based-Compression KD method that outperforms all investigated mainstream KD frameworks on DNS task.

eess.AS↗

arXiv2023

Few-shot Partial Multi-view Learning

It is often the case that data are with multiple views in real-world applications. Fully exploring the information of each view is significant for making data more representative. However, due to various limitations and failures in data collection and pre-processing, it is inevitable for real data to suffer from view missing and data scarcity. The coexistence of these two issues makes it more challenging to achieve the pattern classification task. Currently, to our best knowledge, few appropriate methods can well-handle these two issues simultaneously. Aiming to draw more attention from the community to this challenge, we propose a new task in this paper, called few-shot partial multi-view learning, which focuses on overcoming the negative impact of the view-missing issue in the low-data regime. The challenges of this task are twofold: (i) it is difficult to overcome the impact of data scarcity under the interference of missing views; (ii) the limited number of data exacerbates information scarcity, thus making it harder to address the view-missing issue in turn. To address these challenges, we propose a new unified Gaussian dense-anchoring method. The unified dense anchors are learned for the limited partial multi-view data, thereby anchoring them into a unified dense representation space where the influence of data scarcity and view missing can be alleviated. We conduct extensive experiments to evaluate our method. The results on Cub-googlenet-doc2vec, Handwritten, Caltech102, Scene15, Animal, ORL, tieredImagenet, and Birds-200-2011 datasets validate its effectiveness.

cs.CV↗

arXiv2023

Advancing Incremental Few-shot Semantic Segmentation via Semantic-guided Relation Alignment and Adaptation

Incremental few-shot semantic segmentation (IFSS) aims to incrementally extend a semantic segmentation model to novel classes according to only a few pixel-level annotated data, while preserving its segmentation capability on previously learned base categories. This task faces a severe semantic-aliasing issue between base and novel classes due to data imbalance, which makes segmentation results unsatisfactory. To alleviate this issue, we propose the Semantic-guided Relation Alignment and Adaptation (SRAA) method that fully considers the guidance of prior semantic information. Specifically, we first conduct Semantic Relation Alignment (SRA) in the base step, so as to semantically align base class representations to their semantics. As a result, the embeddings of base classes are constrained to have relatively low semantic correlations to categories that are different from them. Afterwards, based on the semantically aligned base categories, Semantic-Guided Adaptation (SGA) is employed during the incremental learning stage. It aims to ensure affinities between visual and semantic embeddings of encountered novel categories, thereby making the feature representations be consistent with their semantic information. In this way, the semantic-aliasing issue can be suppressed. We evaluate our model on the PASCAL VOC 2012 and the COCO dataset. The experimental results on both these two datasets exhibit its competitive performance, which demonstrates the superiority of our method.

cs.CV↗

arXiv2023

Potential-tuned magnetic switches and half-metallicity transition in zigzag graphene nanoribbons

Realizing controllable room-temperature ferromagnetism in carbon-based materials is one of recent prospects. The magnetism in graphene nanostructures reported previously is mostly formed near the vacancies, zigzag edges, or impurities by breaking the local sublattice imbalance, though a bulk chiral spin-density-wave ground state is also reported at van Hove filling due to its perfectly nested Fermi surface. Here, combining of the first-principles and tight-binding model simulations, we predict a robust ferromagnetic domain lies between the inter-chain carbon atoms inside the zigzag graphene nanoribbons by applying a potential drop. We show that the effective zigzag edges provide the strong correlation background through narrowing the band width, while the internal Van Hove filling provides the strong ferromagnetic background inherited from the bulk. The induced ferromagnetism exhibit interesting switching effect when the nominal Van Hove filling crosses the intra- and inter-chain region by tuning the potential drops. We further observe a robust half-metallicity transition from one spin channel to another within the same magnetic phase. These novel properties provide promising ways to manipulate the spin degree of freedom in graphene nanostructures.

cond-mat.mes-hall↗

arXiv2023

Existence of hyperbolic motions to a class of Hamiltonians and generalized $N$-body system via a geometric approach

For the classical $N$-body problem in $\mathbb{R}^d$ with $d\ge2$, Maderna-Venturelli in their remarkable paper [Ann. Math. 2020] proved the existence of hyperbolic motions with any positive energy constant, starting from any configuration and along any non-collision configuration. Their original proof relies on the long time behavior of solutions by Chazy 1922 and Marchal-Saari 1976, on the Hölder estimate for Mañé's potential by Maderna 2012, and on the weak KAM theory. We give a new and completely different proof for the above existence of hyperbolic motions. The central idea is that, via some geometric observation, we build up uniform estimates for Euclidean length and angle of geodesics of Mañé's potential starting from a given configuration and ending at the ray along a given non-collision configuration. Note that we do not need any of the above previous studies used in Maderna-Venturelli's proof. Moreover, our geometric approach works for Hamiltonians $\frac12\|p\|^2-F(x)$, where $F(x)\ge 0$ is lower semicontinuous and decreases very slowly to $0$ faraway from collisions. We therefore obtain the existence of hyperbolic motions to such Hamiltonians with any positive energy constant, starting from any admissible configuration and along any non-collision configuration. Consequently, for several important potentials $F\in C^{2}(Ω)$, we get similar existence of hyperbolic motions to the generalized $N$-body system $\ddot{x} = \nabla_x F(x)$, which is an extension of Maderna-Venturelli [Ann. Math. 2020].

math.DS↗

arXiv2023

Solid angle measure of polyhedral cones

This paper addresses the computation of normalized solid angle measure of polyhedral cones. This is well understood in dimensions two and three. For higher dimensions, assuming that a positive-definite criterion is met, the measure can be computed via a multivariable hypergeometric series. We present two decompositions of full-dimensional simplicial cones into finite families of cones satisfying the positive-definite criterion, enabling the use of the hypergeometric series to compute the solid angle measure of any polyhedral cone. Additionally, our second decomposition method yields cones with a special tridiagonal structure, reducing the number of required coordinates for the hypergeometric series formula. Furthermore, we investigate the convergence of the hypergeometric series for this case. Our findings provide a powerful tool for computing solid angle measures in high-dimensional spaces.

math.MG↗

arXiv2023

Flopping-mode spin qubit in a Si-MOS quantum dot

Spin qubits based on silicon metal-oxide semiconductor (Si-MOS) quantum dots (QDs) are promising platforms for large-scale quantum computers. To control spin qubits in QDs, electric dipole spin resonance (EDSR) has been most commonly used in recent years. By delocalizing an electron across a double quantum dots charge state, flopping-mode EDSR has been realized in Si/SiGe QDs. Here, we demonstrate a flopping-mode spin qubit in a Si-MOS QD via Elzerman single-shot readout. When changing the detuning with a fixed drive power, we achieve s-shape spin resonance frequencies, an order of magnitude improvement in the spin Rabi frequencies, and virtually constant spin dephasing times. Our results offer a route to large-scale spin qubit systems with higher control fidelity in Si-MOS QDs.

cond-mat.mes-hall↗

arXiv2023

Full tunability and quantum coherent dynamics of a driven multilevel system

Tunability of an artificial quantum system is crucial to its capability to process quantum information. However, tunability usually poses significant demand on the design and fabrication of a device. In this work, we demonstrate that Floquet engineering based on longitudinal driving provides distinct possibilities in enhancing the tunability of a quantum system without needing additional resources. In particular, we study a multilevel model based on gate-defined double quantum dots, where coherent interference occurs when the system is driven longitudinally. We develop an effective model to describe the driven dynamics of this multilevel system, and show that it is highly tunable via the driving field. We then illustrate the versatility and rich physics of a driven multilevel system by exploring phenomena such as driving modulation of resonances, adiabatic state transfer, and dark state. In the context of qubit control, we propose noise-resistant quantum gates based on adiabatic passage. The theoretical consideration we present here is rather general, and is in principle valid for other multilevel quantum systems.

cond-mat.mes-hall↗

arXiv2023

RAF: Holistic Compilation for Deep Learning Model Training

As deep learning is pervasive in modern applications, many deep learning frameworks are presented for deep learning practitioners to develop and train DNN models rapidly. Meanwhile, as training large deep learning models becomes a trend in recent years, the training throughput and memory footprint are getting crucial. Accordingly, optimizing training workloads with compiler optimizations is inevitable and getting more and more attentions. However, existing deep learning compilers (DLCs) mainly target inference and do not incorporate holistic optimizations, such as automatic differentiation and automatic mixed precision, in training workloads. In this paper, we present RAF, a deep learning compiler for training. Unlike existing DLCs, RAF accepts a forward model and in-house generates a training graph. Accordingly, RAF is able to systematically consolidate graph optimizations for performance, memory and distributed training. In addition, to catch up to the state-of-the-art performance with hand-crafted kernel libraries as well as tensor compilers, RAF proposes an operator dialect mechanism to seamlessly integrate all possible kernel implementations. We demonstrate that by in-house training graph generation and operator dialect mechanism, we are able to perform holistic optimizations and achieve either better training throughput or larger batch size against PyTorch (eager and torchscript mode), XLA, and DeepSpeed for popular transformer models on GPUs.

cs.LG↗

arXiv2023

Time-Variance Aware Real-Time Speech Enhancement

Time-variant factors often occur in real-world full-duplex communication applications. Some of them are caused by the complex environment such as non-stationary environmental noises and varying acoustic path while some are caused by the communication system such as the dynamic delay between the far-end and near-end signals. Current end-to-end deep neural network (DNN) based methods usually model the time-variant components implicitly and can hardly handle the unpredictable time-variance in real-time speech enhancement. To explicitly capture the time-variant components, we propose a dynamic kernel generation (DKG) module that can be introduced as a learnable plug-in to a DNN-based end-to-end pipeline. Specifically, the DKG module generates a convolutional kernel regarding to each input audio frame, so that the DNN model is able to dynamically adjust its weights according to the input signal during inference. Experimental results verify that DKG module improves the performance of the model under time-variant scenarios, in the joint acoustic echo cancellation (AEC) and deep noise suppression (DNS) tasks.

eess.AS↗

arXiv2023

Medical visual question answering using joint self-supervised learning

Visual Question Answering (VQA) becomes one of the most active research problems in the medical imaging domain. A well-known VQA challenge is the intrinsic diversity between the image and text modalities, and in the medical VQA task, there is another critical problem relying on the limited size of labelled image-question-answer data. In this study we propose an encoder-decoder framework that leverages the image-text joint representation learned from large-scaled medical image-caption data and adapted to the small-sized medical VQA task. The encoder embeds across the image-text dual modalities with self-attention mechanism and is independently pre-trained on the large-scaled medical image-caption dataset by multiple self-supervised learning tasks. Then the decoder is connected to the top of the encoder and fine-tuned using the small-sized medical VQA dataset. The experiment results present that our proposed method achieves better performance comparing with the baseline and SOTA methods.

cs.CV↗

arXiv2023

Real-time speech enhancement with dynamic attention span

For real-time speech enhancement (SE) including noise suppression, dereverberation and acoustic echo cancellation, the time-variance of the audio signals becomes a severe challenge. The causality and memory usage limit that only the historical information can be used for the system to capture the time-variant characteristics. We propose to adaptively change the receptive field according to the input signal in deep neural network based SE model. Specifically, in an encoder-decoder framework, a dynamic attention span mechanism is introduced to all the attention modules for controlling the size of historical content used for processing the current frame. Experimental results verify that this dynamic mechanism can better track time-variant factors and capture speech-related characteristics, benefiting to both interference removing and speech quality retaining.

eess.AS↗