BZPEERPapers, people, evidence and research paths
Menu

Discover

SearchFind papers, people, topics, institutions and opportunities from one place.GraphSee how papers, authors, topics and evidence connect.

Research tools

HomeContinue from the papers, people and topics that matter now.CollectionsTurn saved papers into reading lists, evidence maps and shareable context.ResearchReconstruct the path that led you from search to paper to question.CollaborationMove from a paper, topic or expert request into a focused thread.PublishCreate full-text research articles with metadata, formulas and versions.

Network

ExpertsFind specialists whose work and reviews explain why they can help.CommunitiesJoin research groups organized around papers, topics and shared evidence.

Opportunities

ListingsBrowse roles, calls and collaborations connected to real research context.

Account

Log inReturn to your saved papers, graph views and active threads.Create accountStart saving papers, building a research trail and joining teams.
Log inCreate account

Source author record

Lu Yu

Lu Yu appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record
cond-mat.str-elComputer VisionMachine Learningcond-matInformation Retrievalcond-mat.supr-conArtificial Intelligencemath.OCNetworking and Internet ArchitectureApplicationsComputationCryptography and SecurityDigital LibrariesDistributed, Parallel, and Cluster Computingeess.IVmath.PRmath.STMethodologyMultimediaphysics.flu-dynphysics.soc-phQuantitative MethodsSocial and Information NetworksStatistics Theory

Catalog footprint

What is connected

34works
24topics
4close collaborators

Actions

Connect this record

Log in to claim
Open graphBrowse works

Research graph

See the researcher in context

Open full explorer

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

34 published item(s)

preprint2026arXiv

On the Limits of Latent Reuse in Diffusion Models

Diffusion models are often trained in low-dimensional latent spaces, which are then reused for related but shifted datasets. In this work, we study when such latent reuse remains reliable under distribution shift. We consider a source-target setting in which both datasets are approximately low-dimensional but may lie near different subspaces. We show that freezing and reusing a source latent space induces a target-domain score error governed by two quantities: the principal-angle misalignment between the source and target subspaces, and the target ambient noise amplified by the diffusion time scale. Motivated by these limits, we further study mixed source-target training and characterize how the required shared latent dimension depends on the relative geometry of the two distributions. Our results provide theoretical guidance on when latent reuse is reliable and when learning a shared representation may be necessary.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Adversarial Robustness of Visual Dialog

Adversarial robustness evaluates the worst-case performance scenario of a machine learning model to ensure its safety and reliability. This study is the first to investigate the robustness of visually grounded dialog models towards textual attacks. These attacks represent a worst-case scenario where the input question contains a synonym which causes the previously correct model to return a wrong answer. Using this scenario, we first aim to understand how multimodal input components contribute to model robustness. Our results show that models which encode dialog history are more robust, and when launching an attack on history, model prediction becomes more uncertain. This is in contrast to prior work which finds that dialog history is negligible for model performance on this task. We also evaluate how to generate adversarial test examples which successfully fool the model but remain undetected by the user/software designer. We find that the textual, as well as the visual context are important to generate plausible worst-case scenarios.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Continually Learning Self-Supervised Representations with Projected Functional Regularization

Recent self-supervised learning methods are able to learn high-quality image representations and are closing the gap with supervised approaches. However, these methods are unable to acquire new knowledge incrementally -- they are, in fact, mostly used only as a pre-training phase over IID data. In this work we investigate self-supervised methods in continual learning regimes without any replay mechanism. We show that naive functional regularization, also known as feature distillation, leads to lower plasticity and limits continual learning performance. Instead, we propose Projected Functional Regularization in which a separate temporal projection network ensures that the newly learned feature space preserves information of the previous one, while at the same time allowing for the learning of new features. This prevents forgetting while maintaining the plasticity of the learner. Comparison with other incremental learning approaches applied to self-supervision demonstrates that our method obtains competitive performance in different scenarios and on multiple datasets.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

eX-ViT: A Novel eXplainable Vision Transformer for Weakly Supervised Semantic Segmentation

Recently vision transformer models have become prominent models for a range of vision tasks. These models, however, are usually opaque with weak feature interpretability. Moreover, there is no method currently built for an intrinsically interpretable transformer, which is able to explain its reasoning process and provide a faithful explanation. To close these crucial gaps, we propose a novel vision transformer dubbed the eXplainable Vision Transformer (eX-ViT), an intrinsically interpretable transformer model that is able to jointly discover robust interpretable features and perform the prediction. Specifically, eX-ViT is composed of the Explainable Multi-Head Attention (E-MHA) module, the Attribute-guided Explainer (AttE) module and the self-supervised attribute-guided loss. The E-MHA tailors explainable attention weights that are able to learn semantically interpretable representations from local patches in terms of model decisions with noise robustness. Meanwhile, AttE is proposed to encode discriminative attribute features for the target object through diverse attribute discovery, which constitutes faithful evidence for the model's predictions. In addition, a self-supervised attribute-guided loss is developed for our eX-ViT, which aims at learning enhanced representations through the attribute discriminability mechanism and attribute diversity mechanism, to localize diverse and discriminative attributes and generate more robust explanations. As a result, we can uncover faithful and robust interpretations with diverse attributes through the proposed eX-ViT.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Mirror Descent Strikes Again: Optimal Stochastic Convex Optimization under Infinite Noise Variance

We study stochastic convex optimization under infinite noise variance. Specifically, when the stochastic gradient is unbiased and has uniformly bounded $(1+κ)$-th moment, for some $κ\in (0,1]$, we quantify the convergence rate of the Stochastic Mirror Descent algorithm with a particular class of uniformly convex mirror maps, in terms of the number of iterations, dimensionality and related geometric parameters of the optimization problem. Interestingly this algorithm does not require any explicit gradient clipping or normalization, which have been extensively used in several recent empirical and theoretical works. We complement our convergence results with information-theoretic lower bounds showing that no other algorithm using only stochastic first-order oracles can achieve improved rates. Our results have several interesting consequences for devising online/streaming stochastic approximation algorithms for problems arising in robust statistics and machine learning.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Q-LIC: Quantizing Learned Image Compression with Channel Splitting

Learned image compression (LIC) has reached a comparable coding gain with traditional hand-crafted methods such as VVC intra. However, the large network complexity prohibits the usage of LIC on resource-limited embedded systems. Network quantization is an efficient way to reduce the network burden. This paper presents a quantized LIC (QLIC) by channel splitting. First, we explore that the influence of quantization error to the reconstruction error is different for various channels. Second, we split the channels whose quantization has larger influence to the reconstruction error. After the splitting, the dynamic range of channels is reduced so that the quantization error can be reduced. Finally, we prune several channels to keep the number of overall channels as origin. By using the proposal, in the case of 8-bit quantization for weight and activation of both main and hyper path, we can reduce the BD-rate by 0.61%-4.74% compared with the previous QLIC. Besides, we can reach better coding gain compared with the state-of-the-art network quantization method when quantizing MS-SSIM models. Moreover, our proposal can be combined with other network quantization methods to further improve the coding gain. The moderate coding loss caused by the quantization validates the feasibility of the hardware implementation for QLIC in the future.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Self-Training for Class-Incremental Semantic Segmentation

In class-incremental semantic segmentation, we have no access to the labeled data of previous tasks. Therefore, when incrementally learning new classes, deep neural networks suffer from catastrophic forgetting of previously learned knowledge. To address this problem, we propose to apply a self-training approach that leverages unlabeled data, which is used for the rehearsal of previous knowledge. Specifically, we first learn a temporary model for the current task, and then pseudo labels for the unlabeled data are computed by fusing information from the old model of the previous task and the current temporary model. Additionally, conflict reduction is proposed to resolve the conflicts of pseudo labels generated from both the old and temporary models. We show that maximizing self-entropy can further improve results by smoothing the overconfident predictions. Interestingly, in the experiments we show that the auxiliary data can be different from the training data and that even general-purpose but diverse auxiliary data can lead to large performance gains. The experiments demonstrate state-of-the-art results: obtaining a relative gain of up to 114% on Pascal-VOC 2012 and 8.5% on the more challenging ADE20K compared to previous state-of-the-art methods.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

False Discovery Rates in Biological Networks

The increasing availability of data has generated unprecedented prospects for network analyses in many biological fields, such as neuroscience (e.g., brain networks), genomics (e.g., gene-gene interaction networks), and ecology (e.g., species interaction networks). A powerful statistical framework for estimating such networks is Gaussian graphical models, but standard estimators for the corresponding graphs are prone to large numbers of false discoveries. In this paper, we introduce a novel graph estimator based on knockoffs that imitate the partial correlation structures of unconnected nodes. We show that this new estimator guarantees accurate control of the false discovery rate in theory, simulations, and biological applications, and we provide easy-to-use R code.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Addressing Class-Imbalance Problem in Personalized Ranking

Pairwise ranking models have been widely used to address recommendation problems. The basic idea is to learn the rank of users' preferred items through separating items into \emph{positive} samples if user-item interactions exist, and \emph{negative} samples otherwise. Due to the limited number of observable interactions, pairwise ranking models face serious \emph{class-imbalance} issues. Our theoretical analysis shows that current sampling-based methods cause the vertex-level imbalance problem, which makes the norm of learned item embeddings towards infinite after a certain training iterations, and consequently results in vanishing gradient and affects the model inference results. We thus propose an efficient \emph{\underline{Vi}tal \underline{N}egative \underline{S}ampler} (VINS) to alleviate the class-imbalance issue for pairwise ranking model, in particular for deep learning models optimized by gradient methods. The core of VINS is a bias sampler with reject probability that will tend to accept a negative candidate with a larger degree weight than the given positive item. Evaluation results on several real datasets demonstrate that the proposed sampling method speeds up the training procedure 30\% to 50\% for ranking models ranging from shallow to deep, while maintaining and even improving the quality of ranking results in top-N item recommendation.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

An Analysis of Constant Step Size SGD in the Non-convex Regime: Asymptotic Normality and Bias

Structured non-convex learning problems, for which critical points have favorable statistical properties, arise frequently in statistical machine learning. Algorithmic convergence and statistical estimation rates are well-understood for such problems. However, quantifying the uncertainty associated with the underlying training algorithm is not well-studied in the non-convex setting. In order to address this shortcoming, in this work, we establish an asymptotic normality result for the constant step size stochastic gradient descent (SGD) algorithm--a widely used algorithm in practice. Specifically, based on the relationship between SGD and Markov Chains [DDB19], we show that the average of SGD iterates is asymptotically normally distributed around the expected value of their unique invariant distribution, as long as the non-convex and non-smooth objective function satisfies a dissipativity property. We also characterize the bias between this expected value and the critical points of the objective function under various local regularity conditions. Together, the above two results could be leveraged to construct confidence intervals for non-convex problems that are trained using the SGD algorithm.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Compressing Facial Makeup Transfer Networks by Collaborative Distillation and Kernel Decomposition

Although the facial makeup transfer network has achieved high-quality performance in generating perceptually pleasing makeup images, its capability is still restricted by the massive computation and storage of the network architecture. We address this issue by compressing facial makeup transfer networks with collaborative distillation and kernel decomposition. The main idea of collaborative distillation is underpinned by a finding that the encoder-decoder pairs construct an exclusive collaborative relationship, which is regarded as a new kind of knowledge for low-level vision tasks. For kernel decomposition, we apply the depth-wise separation of convolutional kernels to build a light-weighted Convolutional Neural Network (CNN) from the original network. Extensive experiments show the effectiveness of the compression method when applied to the state-of-the-art facial makeup transfer network -- BeautyGAN.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Leidenfrost drop impact on inclined superheated substrates

In real applications, drops always impact on solid walls with various inclinations. For the oblique impact of a Leidenfrost drop, which has a vapor layer under its bottom surface to prevent its direct contact with the superheated substrate, the drop can nearly frictionlessly slide along the substrate accompanied by the spreading and the retracting. To individually study these processes, we experimentally observe ethanol drops impact on superheated inclined substrates using high-speed imaging from two different views synchronously. We first study the dynamic Leidenfrost temperature, which mainly depends on the normal Weber number ${We}_\perp $. Then the substrate temperature is set to be high enough to study the Leidenfrost drop behaviors. During the spreading process, drops always keep uniform. And the maximum spreading factor $D_m/D_0$ follows a power-law dependence on the large normal Weber number ${We}_\perp $ as $D_m/D_0 = \sqrt{We_\perp /12+2}$ for $We_\perp \geq 30$. During the retracting process, drops with low impact velocities become non-uniform due to the gravity effect. For the sliding process, the residence time of all studied drops is nearly a constant, which is not affected by the inclination and $We$ number. The frictionless vapor layer results in the dimensionless sliding distance $L/D_0$ follows a power-law dependence on the parallel Weber number $We_\parallel$ as $L/D_0 \propto We_\parallel^{1/2}$. Without direct contact with the substrate, the behaviors of drops can be separately determined by ${We}_\perp $ and $We_\parallel$. When the impact velocity is too high, the drop fragments into many tiny droplets, which is called the splashing phenomenon. The critical splashing criterion is found to be $We_\perp ^*\simeq$ 120 or $K_\perp = We_\perp Re_\perp^{1/2} \simeq$ 5300 in the current parameter regime.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Protocol Proxy: An FTE-based Covert Channel

In a hostile network environment, users must communicate without being detected. This involves blending in with the existing traffic. In some cases, a higher degree of secrecy is required. We present a proof-of-concept format transforming encryption (FTE)-based covert channel for tunneling TCP traffic through protected static protocols. Protected static protocols are UDP-based protocols with variable fields that cannot be blocked without collateral damage, such as power grid failures. We (1) convert TCP traffic to UDP traffic, (2) introduce observation-based FTE, and (3) model interpacket timing with a deterministic Hidden Markov Model (HMM). The resulting Protocol Proxy has a very low probability of detection and is an alternative to current covert channels. We tunnel a TCP session through a UDP protocol and guarantee delivery. Observation-based FTE ensures traffic cannot be detected by traditional rule-based analysis or DPI. A deterministic HMM ensures the Protocol Proxy accurately models interpacket timing to avoid detection by side-channel analysis. Finally, the choice of a protected static protocol foils stateful protocol analysis and causes collateral damage with false positives.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Semantic Drift Compensation for Class-Incremental Learning

Class-incremental learning of deep networks sequentially increases the number of classes to be classified. During training, the network has only access to data of one task at a time, where each task contains several classes. In this setting, networks suffer from catastrophic forgetting which refers to the drastic drop in performance on previous tasks. The vast majority of methods have studied this scenario for classification networks, where for each new task the classification layer of the network must be augmented with additional weights to make room for the newly added classes. Embedding networks have the advantage that new classes can be naturally included into the network without adding new weights. Therefore, we study incremental learning for embedding networks. In addition, we propose a new method to estimate the drift, called semantic drift, of features and compensate for it without the need of any exemplars. We approximate the drift of previous tasks based on the drift that is experienced by current task data. We perform experiments on fine-grained datasets, CIFAR100 and ImageNet-Subset. We demonstrate that embedding networks suffer significantly less from catastrophic forgetting. We outperform existing methods which do not require exemplars and obtain competitive results compared to methods which store exemplars. Furthermore, we show that our proposed SDC when combined with existing methods to prevent forgetting consistently improves results.

0 citations0 reviewsNo code linkedOpen access
preprint2016arXiv

Identifying the Academic Rising Stars

Predicting the fast-rising young researchers (Academic Rising Stars) in the future provides useful guidance to the research community, e.g., offering competitive candidates to university for young faculty hiring as they are expected to have success academic careers. In this work, given a set of young researchers who have published the first first-author paper recently, we solve the problem of how to effectively predict the top k% researchers who achieve the highest citation increment in Δt years. We explore a series of factors that can drive an author to be fast-rising and design a novel impact increment ranking learning (IIRL) algorithm that leverages those factors to predict the academic rising stars. Experimental results on the large ArnetMiner dataset with over 1.7 million authors demonstrate the effectiveness of IIRL. Specifically, it outperforms all given benchmark methods, with over 8% average improvement. Further analysis demonstrates that the prediction models for different research topics follow the similar pattern. We also find that temporal features are the best indicators for rising stars prediction, while venue features are less relevant.

0 citations0 reviewsNo code linkedOpen access
preprint2016arXiv

Pairing symmetry of heavy fermion superconductivity in the two-dimensional Kondo-Heisenberg lattice model

In the two-dimensional Kondo-Heisenberg lattice model away from half-filled, the local antiferromagnetic exchange coupling can provide the pairing mechanism of quasiparticles via the Kondo screening effect, leading to the heavy fermion superconductivity. We find that the pairing symmetry \textit{strongly} depends on the Fermi surface (FS) structure in the normal metallic state. When $J_{H}/J_{K}$ is very small, the FS is a small hole-like circle around the corner of the Brillouin zone, and the s-wave pairing symmetry has a lower ground state energy. For the intermediate coupling values of $J_{H}/J_{K}$, the extended s-wave pairing symmetry gives the favored ground state. However, when $J_{H}/J_{K}$ is larger than a critical value, the FS transforms into four small hole pockets crossing the boundary of the magnetic Brillouin zone, and the d-wave pairing symmetry becomes more favorable. In that regime, the resulting superconducting state is characterized by either nodal d-wave or nodeless d-wave state, depending on the conduction electron filling factor as well. A continuous phase transition exists between these two states. This result may be related to the phase transition of the nodal d-wave state to a fully gapped state, which is recently observed in Yb doped CeCoIn$_{5}$.

0 citations0 reviewsNo code linkedOpen access
preprint2015arXiv

Efficient Channel-Hopping Rendezvous Algorithm Based on Available Channel Set

In cognitive radio networks, rendezvous is a fundamental operation by which two cognitive users establish a communication link on a commonly-available channel for communications. Some existing rendezvous algorithms can guarantee that rendezvous can be completed within finite time and they generate channel-hopping (CH) sequences based on the whole channel set. However, some channels may not be available (e.g., they are being used by the licensed users) and these existing algorithms would randomly replace the unavailable channels in the CH sequence. This random replacement is not effective, especially when the number of unavailable channels is large. In this paper, we design a new rendezvous algorithm that attempts rendezvous on the available channels only for faster rendezvous. This new algorithm, called Interleaved Sequences based on Available Channel set (ISAC), constructs an odd sub-sequence and an even sub-sequence and interleaves these two sub-sequences to compose a CH sequence. We prove that ISAC provides guaranteed rendezvous (i.e., rendezvous can be achieved within finite time). We derive the upper bound on the maximum time-to-rendezvous (MTTR) to be O(m) (m is not greater than Q) under the symmetric model and O(mn) (n is not greater than Q) under the asymmetric model, where m and n are the number of available channels of two users and Q is the total number of channels (i.e., all potentially available channels). We conduct extensive computer simulation to demonstrate that ISAC gives significantly smaller MTTR than the existing algorithms.

0 citations0 reviewsNo code linkedOpen access
preprint2015arXiv

Multi-Linear Interactive Matrix Factorization

Recommender systems, which can significantly help users find their interested items from the information era, has attracted an increasing attention from both the scientific and application society. One of the widest applied recommendation methods is the Matrix Factorization (MF). However, most of MF based approaches focus on the user-item rating matrix, but ignoring the ingredients which may have significant influence on users' preferences on items. In this paper, we propose a multi-linear interactive MF algorithm (MLIMF) to model the interactions between the users and each event associated with their final decisions. Our model considers not only the user-item rating information but also the pairwise interactions based on some empirically supported factors. In addition, we compared the proposed model with three typical other methods: user-based collaborative filtering (UCF), item-based collaborative filtering (ICF) and regularized MF (RMF). Experimental results on two real-world datasets, \emph{MovieLens} 1M and \emph{MovieLens} 100k, show that our method performs much better than other three methods in the accuracy of recommendation. This work may shed some light on the in-depth understanding of modeling user online behaviors and the consequent decisions.

0 citations0 reviewsNo code linkedOpen access
preprint2015arXiv

Phase evolution of the two-dimensional Kondo lattice model near half-filling

Within a mean-field approximation, the ground state and finite temperature phase diagrams of the two-dimensional Kondo lattice model have been carefully studied as functions of the Kondo coupling $J$ and the conduction electron concentration $n_{c}$. In addition to the conventional hybridization between local moments and itinerant electrons, a staggered hybridization is proposed to characterize the interplay between the antiferromagnetism and the Kondo screening effect. As a result, a heavy fermion antiferromagnetic phase is obtained and separated from the pure antiferromagnetic ordered phase by a first-order Lifshitz phase transition, while a continuous phase transition exists between the heavy fermion antiferromagnetic phase and the Kondo paramagnetic phase. We have developed a efficient theory to calculate these phase boundaries. As $n_{c}$ decreases from the half-filling, the region of the heavy fermion antiferromagnetic phase shrinks and finally disappears at a critical point $n_{c}^{*}=0.8228$, leaving a first-order critical line between the pure antiferromagnetic phase and the Kondo paramagnetic phase for $n_{c}<n_{c}^{* }$. At half-filling limit, a finite temperature phase diagram is also determined on the Kondo coupling and temperature ($J$-$T$) plane. Notably, as the temperature is increased, the region of the heavy fermion antiferromagnetic phase is reduced continuously, and finally converges to a single point, together with the pure antiferromagnetic phase and the Kondo paramagnetic phase. The phase diagrams with such triple point may account for the observed phase transitions in related heavy fermion materials.

0 citations0 reviewsNo code linkedOpen access
preprint2015arXiv

ZOS: A Fast Rendezvous Algorithm Based on Set of Available Channels for Cognitive Radios

Most of existing rendezvous algorithms generate channel-hopping sequences based on the whole channel set. They are inefficient when the set of available channels is a small subset of the whole channel set. We propose a new algorithm called ZOS which uses three types of elementary sequences (namely, Zero-type, One-type, and S-type) to generate channel-hopping sequences based on the set of available channels. ZOS provides guaranteed rendezvous without any additional requirements. The maximum time-to-rendezvous of ZOS is upper-bounded by O(m1*m2*log2M) where M is the number of all channels and m1 and m2 are the numbers of available channels of two users.

0 citations0 reviewsNo code linkedOpen access
preprint2014arXiv

ILCR: Item-based Latent Factors for Sparse Collaborative Retrieval

Interactions between search and recommendation have recently attracted significant attention, and several studies have shown that many potential applications involve with a joint problem of producing recommendations to users with respect to a given query, termed $Collaborative$ $Retrieval$ (CR). Successful algorithms designed for CR should be potentially flexible at dealing with the sparsity challenges since the setup of collaborative retrieval associates with a given $query$ $\times$ $user$ $\times$ $item$ tensor instead of traditional $user$ $\times$ $item$ matrix. Recently, several works are proposed to study CR task from users' perspective. In this paper, we aim to sufficiently explore the sophisticated relationship of each $query$ $\times$ $user$ $\times$ $item$ triple from items' perspective. By integrating item-based collaborative information for this joint task, we present an alternative factorized model that could better evaluate the ranks of those items with sparse information for the given query-user pair. In addition, we suggest to employ a recently proposed scalable ranking learning algorithm, namely BPR, to optimize the state-of-the-art approach, $Latent$ $Collaborative$ $Retrieval$ model, instead of the original learning algorithm. The experimental results on two real-world datasets, (i.e. \emph{Last.fm}, \emph{Yelp}), demonstrate the efficiency and effectiveness of our proposed approach.

0 citations0 reviewsNo code linkedOpen access
preprint2014arXiv

Information Filtering via Collaborative User Clustering Modeling

The past few years have witnessed the great success of recommender systems, which can significantly help users find out personalized items for them from the information era. One of the most widely applied recommendation methods is the Matrix Factorization (MF). However, most of researches on this topic have focused on mining the direct relationships between users and items. In this paper, we optimize the standard MF by integrating the user clustering regularization term. Our model considers not only the user-item rating information, but also takes into account the user interest. We compared the proposed model with three typical other methods: User-Mean (UM), Item-Mean (IM) and standard MF. Experimental results on a real-world dataset, \emph{MovieLens}, show that our method performs much better than other three methods in the accuracy of recommendation.

0 citations0 reviewsNo code linkedOpen access
preprint2013arXiv

Weak ferromagnetism with the Kondo screening effect in the Kondo lattice systems

We carefully consider the interplay between ferromagnetism and the Kondo screening effect in the conventional Kondo lattice systems at finite temperatures. Within an effective mean-field theory for small conduction electron densities, a complete phase diagram has been determined. In the ferromagnetic ordered phase, there is a characteristic temperature scale to indicate the presence of the Kondo screening effect. We further find two distinct ferromagnetic long-range ordered phases coexisting with the Kondo screening effect: spin fully polarized and partially polarized states. A continuous phase transition exists to separate the partially polarized ferromagnetic ordered phase from the paramagnetic heavy Fermi liquid phase. These results may be used to explain the weak ferromagnetism observed recently in the Kondo lattice materials.

0 citations0 reviewsNo code linkedOpen access
preprint2011arXiv

Accurate determination of the Gaussian transition in spin-1 chains with single-ion anisotropy

The Gaussian transition in the spin-one Heisenberg chain with single-ion anisotropy is extremely difficult to treat, both analytically and numerically. We introduce an improved DMRG procedure with strict error control, which we use to access very large systems. By considering the bulk entropy, we determine the Gaussian transition point to 4-digit accuracy, $D_{c}/J = 0.96845(8)$, resolving a long-standing debate in quantum magnetism. With this value, we obtain high-precision data for the critical behavior of quantities including the ground-state energy, gap, and transverse string-order parameter, and for the critical exponent, $ν= 1.472(2)$. Applying our improved technique at $J_{z} = 0.5$ highlights essential differences in critical behavior along the Gaussian transition line.

0 citations0 reviewsNo code linkedOpen access
preprint2010arXiv

Insulator-to-metal phase transition in Yb-based Kondo insulators

The periodic Anderson lattice model for the crystalline electric field (CEF)split 4f quartet states is used to describe the Yb-based Kondo insulators/semiconductors. In the slave-boson mean-field approximation, we derive the hybridized quasiparticle bands, and find that decreasing the hybridization difference of the two CEF quartets may induce an insulator-to-metal phase transition. The resulting metallic phase has a hole and an electron Fermi pockets. Such a phase transition may be realized experimentally by applying pressure, reducing the difference in hybridization of the two CEF quartets.

0 citations0 reviewsNo code linkedOpen access
preprint2010arXiv

Kondo screening coexisting with ferromagnetic order as a possible ground state for Kondo lattice systems

We consider the competition between the Kondo screening effect and ferromagnetic long-range order (FLRO) within a mean-field theory of the Kondo lattice model for low conduction electron densities $n_{c}$. Depending on the parameter values, several types of FLRO ground states are found. When $n_{c}<0.16$, a polarized FLRO phase is dominant. For $0.16<n_{c}<0.82$, a non-polarized FLRO phase appears in the weak Kondo coupling region; while in the intermediate coupling region the ground state corresponds to the polarized and non-polarized FLRO phases, respectively, coexisting with the Kondo screening. For a strong Kondo coupling, the product of pure Kondo singlets is the ground state. Moreover, we also find that a weak magnetic field makes the pure Kondo singlet phase vanish, while the non-polarized FLRO state with the Kondo screening spans a large area in the phase diagram.

0 citations0 reviewsNo code linkedOpen access
preprint2010arXiv

Lifshitz transitions in a heavy-Fermion liquid driven by short-range antiferromagnetic correlations in the two-dimensional Kondo lattice model

The heavy-Fermion liquid with short-range antiferromagnetic correlations is carefully considered in the two-dimensional Kondo-Heisenberg lattice model. As the ratio of the local Heisenberg superexchange $J_{H}$ to the Kondo coupling $J_{K}$ increases, Lifshitz transitions are anticipated, where the topology of the Fermi surface (FS) of the heavy quasiparticles changes from a hole-like circle to four kidney-like pockets centered around $(π,π)$. In-between these two limiting cases, a first-order quantum phase transition is identified at $J_{H}/J_{K}=0.1055$ where a small circle begins to emerge within the large deformed circle. When $J_{H}/J_{K}=0.1425$, the two deformed circles intersect each other and then decompose into four kidney-like Fermi pockets via a second-order quantum phase transition. As $J_{H}/J_{K}$ increases further, the Fermi pockets are shifted along the direction ($π,π$) to ($π/2,π/2$), and the resulting FS is consistent with the FS obtained recently using the quantum Monte Carlo cluster approach to the Kondo lattice system in the presence of the antiferrmagnetic order.

0 citations0 reviewsNo code linkedOpen access
preprint2005arXiv

Midgap States in Antiferromagnetic Heisenberg Chains with A Staggered Field

We study low-energy excitations in antiferromagnetic Heisenberg chains with a staggered field which splits the spectrum into a longitudinal and a transverse branch. Bound states are found to exist inside the field induced gap in both branches. They originate from the edge effects and are inherent to spin-chain materials. The sine-Gordon scaling $h_s^{2/3}|\log h_s|^{1/6}$ ($h_s$: the staggered field) provides an accurate description for the gap and midgap energies in the transverse branch for $S=1/2$ and the midgap energies in both branches for $S=3/2$ over a wide range of magnetic field; however, it can fit other low-energy excitations only at much lower field. Moreover, the integer-spin S=1 chain displays scaling behavior that does not fit this scaling law. These results reveal intriguing features of magnetic excitations in spin-chain materials that deserve further investigation.

0 citations0 reviewsNo code linkedOpen access
preprint2003arXiv

Spin-orbital gapped phase with least symmetry breaking in the one-dimensional symmetrically coupled spin-orbital model

To describe the spin-orbital energy gap formation in the one-dimensional symmetrically coupled spin-orbital model, we propose a simple mean field theory based on an SU(4) constraint fermion representation of spins and orbitals. A spin-orbital gapped phase is formed due to a marginally relevant spin-orbital valence bond pairing interaction. The energy gap of the spin and orbital excitations grows extremely slowly from the SU(4) symmetric point up to a maximum value and then decreases rapidly. By calculating the spin, orbital, and spin-orbital tensor static susceptibilities at zero temperature, we find a crossover from coherent to incoherent magnetic excitations as the spin-orbital coupling decreasing from large to small values.

0 citations0 reviewsNo code linkedOpen access
preprint2002arXiv

Interplay of quantum magnetic and potential scattering around Zn or Ni impurity ions in superconducting cuprates

To describe the scattering of superconducting quasiparticles from non-magnetic (Zn) or magnetic (Ni) impurities in optimally doped high T$_c$ cuprates, we propose an effective Anderson model Hamiltonian of a localized electron hybridizing with $d_{x^2-y^2}$-wave BCS type superconducting quasiparticles with an attractive scalar potential at the impurity site. Due to the strong local antiferromagnetic couplings between the original Cu ions and their nearest neighbors, the localized electron in the Ni-doped materials is assumed to be on the impurity sites, while in the Zn-doped materials the localized electron is distributed over the four nearest neighbor sites of the impurities with a dominant $d_{x^2-y^2}$ symmetric form of the wave function. With Ni impurities, two resonant states are formed above the Fermi level in the local density of states at the impurity site, while for Zn impurities a sharp resonant peak below the Fermi level dominates in the local density of states at the Zn site, accompanied by a small and broad resonant state above the Fermi level mainly induced by the potential scattering. In both cases, there are no Kondo screening effects. The local density of states and their spatial distribution at the dominant resonant energy around the substituted impurities are calculated for both cases, and they are in good agreement with the experimental results of scanning tunneling microscopy in Bi$_2$Sr$_2$CaCu$_2$O$_{8+δ}$ with Zn or Ni impurities, respectively.

0 citations0 reviewsNo code linkedOpen access
preprint2001arXiv

Mesoscopic Kondo screening effect in a single-electron transistor embedded in a metallic ring

We study the Kondo screening effect generated by a single-electron transistor or quantum dot embedded in a small metallic ring. When the ring circumference $L$ becomes comparable to the fundamental length scale $ξ_K^0=\hbar \upsilon _F/T_K^0$ associated with the {\it bulk} Kondo tempe the Kondo resonance is strongly affected, depending on the total number of electrons ({\it modulo} 4) and magnetic flux threading the ring. The resulting Kondo-assisted persistent currents are also calculated in both Kondo and mixed valence regimes, and the maximum values are found in the crossover region.

0 citations0 reviewsNo code linkedOpen access
preprint2000arXiv

Marginal Fermi liquid resonance induced by a quantum magnetic impurity in d-wave superconductors

We consider a model of an Anderson impurity embedded in a $d_{x^2-y^2}$-wave superconducting state to describe the low-energy excitations of cuprate superconductors doped with a small amount of magnetic impurities. Due to the Dirac-like energy dispersion, a sharp localized resonance above the Fermi energy, showing a marginal Fermi liquid behavior ($ω\ln ω$ as $ω\to 0$) is predicted for the impurity states. The same logarithmic dependence of self-energy and a linear frequency dependence of the relaxation rate are also derived for the conduction electrons, characterizing a new universality class for the strong coupling fixed point. At the resonant energies, the spatial distribution of the electron density of states around the magnetic impurity is calculated, to be confronted with measurements of the scanning tunneling microscopy on Bi$_2$Sr$_2$Ca(Cu$_{1-x}$Ni$_x$)O$_{8+δ}$.

0 citations0 reviewsNo code linkedOpen access
preprint1995arXiv

Impurity Energy Level Within The Haldane Gap

An impurity bond $J{'}$ in a periodic 1D antiferromagnetic, spin 1 chain with exchange $J$ is considered. Using the numerical density matrix renormalization group method, we find an impurity energy level in the Haldane gap, corresponding to a bound state near the impurity bond. When $J{'}<J$ the level changes gradually from the edge of the Haldane gap to the ground state energy as the deviation $dev=(J-J{'})/J$ changes from 0 to 1. It seems that there is no threshold. Yet, there is a threshold when $J{'}>J$. The impurity level appears only when the deviation $dev=(J{'}-J)/J{'}$ is greater than $B_{c}$, which is near 0.3 in our calculation.

0 citations0 reviewsNo code linkedOpen access
preprint1993arXiv

The Haldane Energy Gap of A Doped Linear-Chain Heisenberg Antiferromagnet

Using the valence-bond-solid (VBS) approach and the Schwinger boson mean field approximation, we study the dependence of the Haldane gap of a spin-1 linear chain Heisenberg antiferromagnet on impurity doping with different spins. The impurity spins affect the singlet pairing order parameter $Δ$ and the constraint factor $λ$. As a result, the Haldane gap is reduced by a factor $ \sim n_i^{2/3}$, with $n_i$ as the impurity concentration, and eventually collapses at $n_i \sim 1/ξ$ with $ξ$ as the VBS correlation length. This theoretical prediction can be verified by neutron scattering experiments.

0 citations0 reviewsNo code linkedOpen access