arXiv subjects

Tong Chen

Publications and source records attributed to Tong Chen.

At least 217 records · Page 12Linked to original sources

arXiv2022

Choosing good subsamples for regression modelling

A common problem in health research is that we have a large database with many variables measured on a large number of individuals. We are interested in measuring additional variables on a subsample; these measurements may be newly available, or expensive, or simply not considered when the data were first collected. The intended use for the new measurements is to fit a regression model generalisable to the whole cohort (and to its source population). This is a two-phase sampling problem; it differs from some other two-phase sampling problems in the richness of the phase I data and in the goal of regression modelling. In particular, an important special case is measurement-error models, where a variable strongly correlated with the phase II measurements is available at phase I. We will explain how influence functions have been useful as a unifying concept for extending classical results to this setting, and describe the steps from designing for a simple weighted estimator at known parameter values through adaptive multiwave designs and the use of prior information. We will conclude with some comments on the information gap between design-based and model-based estimators in this setting.

stat.ME↗

arXiv2022

DeHIN: A Decentralized Framework for Embedding Large-scale Heterogeneous Information Networks

Modeling heterogeneity by extraction and exploitation of high-order information from heterogeneous information networks (HINs) has been attracting immense research attention in recent times. Such heterogeneous network embedding (HNE) methods effectively harness the heterogeneity of small-scale HINs. However, in the real world, the size of HINs grow exponentially with the continuous introduction of new nodes and different types of links, making it a billion-scale network. Learning node embeddings on such HINs creates a performance bottleneck for existing HNE methods that are commonly centralized, i.e., complete data and the model are both on a single machine. To address large-scale HNE tasks with strong efficiency and effectiveness guarantee, we present \textit{Decentralized Embedding Framework for Heterogeneous Information Network} (DeHIN) in this paper. In DeHIN, we generate a distributed parallel pipeline that utilizes hypergraphs in order to infuse parallelization into the HNE task. DeHIN presents a context preserving partition mechanism that innovatively formulates a large HIN as a hypergraph, whose hyperedges connect semantically similar nodes. Our framework then adopts a decentralized strategy to efficiently partition HINs by adopting a tree-like pipeline. Then, each resulting subnetwork is assigned to a distributed worker, which employs the deep information maximization theorem to locally learn node embeddings from the partition it receives. We further devise a novel embedding alignment scheme to precisely project independently learned node embeddings from all subnetworks onto a common vector space, thus allowing for downstream tasks like link prediction and node classification.

cs.LG↗

arXiv2021

Spin-excitation anisotropy in the nematic state of detwinned FeSe

The origin of the electronic nematicity in FeSe is one of the most important unresolved puzzles in the study of iron-based superconductors. In both spin- and orbital-nematic models, the intrinsic magnetic excitations at $\mathbf{Q}_1=(1, 0)$ and $\mathbf{Q}_2=(0, 1)$ of twin-free FeSe are expected to provide decisive criteria for clarifying this issue. Although a spin-fluctuation anisotropy below 10 meV between $\mathbf{Q}_1$ and $\mathbf{Q}_2$ has been observed by inelastic neutron scattering around $T_c\sim 9$ K ($<<T_s\sim 90$ K), it remains unclear whether such an anisotropy also persists at higher energies and associates with the nematic transition $T_{\rm s}$. Here we use resonant inelastic x-ray scattering (RIXS) to probe the high-energy magnetic excitations of uniaxial-strain detwinned FeSe and {\BFA}. A prominent anisotropy between the magnetic excitations along the $H$ and $K$ directions is found to persist to $\sim200$ meV in FeSe, which is even more pronounced than the anisotropy of spin waves in {\BFA}. This anisotropy decreases gradually with increasing temperature and finally vanishes at a temperature around the nematic transition temperature $T_{\rm s}$. Our results reveal an unprecedented strong spin-excitation anisotropy with a large energy scale well above the $d_{xz}/d_{yz}$ orbital splitting, suggesting that the nematic phase transition is primarily spin-driven. Moreover, the measured high-energy spin excitations are dispersive and underdamped, which can be understood from a local-moment perspective. Our findings provide the much-needed understanding of the mechanism for the nematicity of FeSe and points to a unified description of the correlation physics across seemingly distinct classes of Fe-based superconductors.

cond-mat.supr-con↗

arXiv2021

Personalized On-Device E-health Analytics with Decentralized Block Coordinate Descent

Actuated by the growing attention to personal healthcare and the pandemic, the popularity of E-health is proliferating. Nowadays, enhancement on medical diagnosis via machine learning models has been highly effective in many aspects of e-health analytics. Nevertheless, in the classic cloud-based/centralized e-health paradigms, all the data will be centrally stored on the server to facilitate model training, which inevitably incurs privacy concerns and high time delay. Distributed solutions like Decentralized Stochastic Gradient Descent (D-SGD) are proposed to provide safe and timely diagnostic results based on personal devices. However, methods like D-SGD are subject to the gradient vanishing issue and usually proceed slowly at the early training stage, thereby impeding the effectiveness and efficiency of training. In addition, existing methods are prone to learning models that are biased towards users with dense data, compromising the fairness when providing E-health analytics for minority groups. In this paper, we propose a Decentralized Block Coordinate Descent (D-BCD) learning framework that can better optimize deep neural network-based models distributed on decentralized devices for E-health analytics. Benchmarking experiments on three real-world datasets illustrate the effectiveness and practicality of our proposed D-BCD, where additional simulation study showcases the strong applicability of D-BCD in real-life E-health scenarios.

cs.LG↗

arXiv2021

Spin Waves and Dirac Magnons in a Honeycomb Lattice Zig-zag Antiferromagnet BaNi2(AsO4)2

The topological properties of massive and massless fermionic quasiparticles have been intensively investigated over the past decade in topological materials without magnetism. Recently, the bosonic analogs of such quasiparticles arising from spin waves have been reported in the two-dimensional (2D) honeycomb lattice ferromagnet/antiferromagnet and the 3D antiferromagnet. Here we use time-of-flight inelastic neutron scattering to study spin waves of the S = 1 honeycomb lattice antiferromagnet BaNi2(AsO4)2, which has a zig-zag antiferromagnetic (AF) ground state identical to that of the Kitaev quantum spin liquid candidate alpha-RuCl3. We determine the magnetic exchange interactions in the zig-zag AF ordered phase, and show that spin waves in BaNi2(AsO4)2 have symmetry-protected Dirac points inside the Brillouin zone boundary. These results provide a microscopic understanding of the zig-zag AF order and associated Dirac magnons in honeycomb lattice magnets, and are also important for establishing the magnetic interactions in Kitaev quantum spin liquid candidates.

cond-mat.str-el↗

arXiv2021

SurvODE: Extrapolating Gene Expression Distribution for Early Cancer Identification

With the increasingly available large-scale cancer genomics datasets, machine learning approaches have played an important role in revealing novel insights into cancer development. Existing methods have shown encouraging performance in identifying genes that are predictive for cancer survival, but are still limited in modeling the distribution over genes. Here, we proposed a novel method that can simulate the gene expression distribution at any given time point, including those that are out of the range of the observed time points. In order to model the irregular time series where each patient is one observation, we integrated a neural ordinary differential equation (neural ODE) with cox regression into our framework. We evaluated our method on eight cancer types on TCGA and observed a substantial improvement over existing approaches. Our visualization results and further analysis indicate how our method can be used to simulate expression at the early cancer stage, offering the possibility for early cancer identification.

q-bio.GN↗

arXiv2021

MetaCURE: Meta Reinforcement Learning with Empowerment-Driven Exploration

Meta reinforcement learning (meta-RL) extracts knowledge from previous tasks and achieves fast adaptation to new tasks. Despite recent progress, efficient exploration in meta-RL remains a key challenge in sparse-reward tasks, as it requires quickly finding informative task-relevant experiences in both meta-training and adaptation. To address this challenge, we explicitly model an exploration policy learning problem for meta-RL, which is separated from exploitation policy learning, and introduce a novel empowerment-driven exploration objective, which aims to maximize information gain for task identification. We derive a corresponding intrinsic reward and develop a new off-policy meta-RL framework, which efficiently learns separate context-aware exploration and exploitation policies by sharing the knowledge of task inference. Experimental evaluation shows that our meta-RL method significantly outperforms state-of-the-art baselines on various sparse-reward MuJoCo locomotion tasks and more complex sparse-reward Meta-World tasks.

cs.AI↗

arXiv2021

PipAttack: Poisoning Federated Recommender Systems forManipulating Item Promotion

Due to the growing privacy concerns, decentralization emerges rapidly in personalized services, especially recommendation. Also, recent studies have shown that centralized models are vulnerable to poisoning attacks, compromising their integrity. In the context of recommender systems, a typical goal of such poisoning attacks is to promote the adversary's target items by interfering with the training dataset and/or process. Hence, a common practice is to subsume recommender systems under the decentralized federated learning paradigm, which enables all user devices to collaboratively learn a global recommender while retaining all the sensitive data locally. Without exposing the full knowledge of the recommender and entire dataset to end-users, such federated recommendation is widely regarded `safe' towards poisoning attacks. In this paper, we present a systematic approach to backdooring federated recommender systems for targeted item promotion. The core tactic is to take advantage of the inherent popularity bias that commonly exists in data-driven recommenders. As popular items are more likely to appear in the recommendation list, our innovatively designed attack model enables the target item to have the characteristics of popular items in the embedding space. Then, by uploading carefully crafted gradients via a small number of malicious users during the model update, we can effectively increase the exposure rate of a target (unpopular) item in the resulted federated recommender. Evaluations on two real-world datasets show that 1) our attack model significantly boosts the exposure rate of the target item in a stealthy way, without harming the accuracy of the poisoned recommender; and 2) existing defenses are not effective enough, highlighting the need for new defenses against our local model poisoning attacks to federated recommender systems.

cs.IR↗

arXiv2021

Localization with Sampling-Argmax

Soft-argmax operation is commonly adopted in detection-based methods to localize the target position in a differentiable manner. However, training the neural network with soft-argmax makes the shape of the probability map unconstrained. Consequently, the model lacks pixel-wise supervision through the map during training, leading to performance degradation. In this work, we propose sampling-argmax, a differentiable training method that imposes implicit constraints to the shape of the probability map by minimizing the expectation of the localization error. To approximate the expectation, we introduce a continuous formulation of the output distribution and develop a differentiable sampling process. The expectation can be approximated by calculating the average error of all samples drawn from the output distribution. We show that sampling-argmax can seamlessly replace the conventional soft-argmax operation on various localization tasks. Comprehensive experiments demonstrate the effectiveness and flexibility of the proposed method. Code is available at https://github.com/Jeff-sjtu/sampling-argmax

cs.CV↗

arXiv2021

Analysis of Error-prone Electronic Health Records with Multi-wave Validation Sampling: Association of Maternal Weight Gain during Pregnancy with Childhood Outcomes

Electronic health record (EHR) data are increasingly used for biomedical research, but these data have recognized data quality challenges. Data validation is necessary to use EHR data with confidence, but limited resources typically make complete data validation impossible. Using EHR data, we illustrate prospective, multi-wave, two-phase validation sampling to estimate the association between maternal weight gain during pregnancy and the risks of her child developing obesity or asthma. The optimal validation sampling design depends on the unknown efficient influence functions of regression coefficients of interest. In the first wave of our multi-wave validation design, we estimate the influence function using the unvalidated (phase 1) data to determine our validation sample; then in subsequent waves, we re-estimate the influence function using validated (phase 2) data and update our sampling. For efficiency, estimation combines obesity and asthma sampling frames while calibrating sampling weights using generalized raking. We validated 996 of 10,335 mother-child EHR dyads in 6 sampling waves. Estimated associations between childhood obesity/asthma and maternal weight gain, as well as other covariates, are compared to naive estimates that only use unvalidated data. In some cases, estimates markedly differ, underscoring the importance of efficient validation sampling to obtain accurate estimates incorporating validated data.

stat.AP↗

arXiv2021

Secure Your Ride: Real-time Matching Success Rate Prediction for Passenger-Driver Pairs

In recent years, online ride-hailing platforms have become an indispensable part of urban transportation. After a passenger is matched up with a driver by the platform, both the passenger and the driver have the freedom to simply accept or cancel a ride with one click. Hence, accurately predicting whether a passenger-driver pair is a good match turns out to be crucial for ride-hailing platforms to devise instant order assignments. However, since the users of ride-hailing platforms consist of two parties, decision-making needs to simultaneously account for the dynamics from both the driver and the passenger sides. This makes it more challenging than traditional online advertising tasks. Moreover, the amount of available data is severely imbalanced across different cities, creating difficulties for training an accurate model for smaller cities with scarce data. Though a sophisticated neural network architecture can help improve the prediction accuracy under data scarcity, the overly complex design will impede the model's capacity of delivering timely predictions in a production environment. In the paper, to accurately predict the MSR of passenger-driver, we propose the Multi-View model (MV) which comprehensively learns the interactions among the dynamic features of the passenger, driver, trip order, as well as context. Regarding the data imbalance problem, we further design the Knowledge Distillation framework (KD) to supplement the model's predictive power for smaller cities using the knowledge from cities with denser data and also generate a simple model to support efficient deployment. Finally, we conduct extensive experiments on real-world datasets from several different cities, which demonstrates the superiority of our solution.

cs.LG↗

arXiv2021

Fast-adapting and Privacy-preserving Federated Recommender System

In the mobile Internet era, the recommender system has become an irreplaceable tool to help users discover useful items, and thus alleviating the information overload problem. Recent deep neural network (DNN)-based recommender system research have made significant progress in improving prediction accuracy, which is largely attributed to the access to a large amount of users' personal data collected from users' devices and then centrally stored in the cloud server. However, as there are rising concerns around the globe on user privacy leakage in the online platform, the public is becoming anxious by such abuse of user privacy. Therefore, it is urgent and beneficial to develop a recommender system that can achieve both high prediction accuracy and high degree of user privacy protection. To this end, we propose a DNN-based recommendation model called PrivRec running on the decentralized federated learning (FL) environment, which ensures that a user's data never leaves his/her during the course of model training. On the other hand, to better embrace the data heterogeneity commonly existing in FL, we innovatively introduce a first-order meta-learning method that enables fast in-device personalization with only few data points. Furthermore, to defense from potential malicious participant that poses serious security threat to other users, we develop a user-level differentially private DP-PrivRec model so that it is unable to determine whether a particular user is present or not solely based on the trained model. Finally, we conduct extensive experiments on two large-scale datasets in a simulated FL environment, and the results validate the superiority of our proposed PrivRec and DP-PrivRec.

cs.IR↗

arXiv2021

Lightweight Self-Attentive Sequential Recommendation

Modern deep neural networks (DNNs) have greatly facilitated the development of sequential recommender systems by achieving state-of-the-art recommendation performance on various sequential recommendation tasks. Given a sequence of interacted items, existing DNN-based sequential recommenders commonly embed each item into a unique vector to support subsequent computations of the user interest. However, due to the potentially large number of items, the over-parameterised item embedding matrix of a sequential recommender has become a memory bottleneck for efficient deployment in resource-constrained environments, e.g., smartphones and other edge devices. Furthermore, we observe that the widely-used multi-head self-attention, though being effective in modelling sequential dependencies among items, heavily relies on redundant attention units to fully capture both global and local item-item transition patterns within a sequence. In this paper, we introduce a novel lightweight self-attentive network (LSAN) for sequential recommendation. To aggressively compress the original embedding matrix, LSAN leverages the notion of compositional embeddings, where each item embedding is composed by merging a group of selected base embedding vectors derived from substantially smaller embedding matrices. Meanwhile, to account for the intrinsic dynamics of each item, we further propose a temporal context-aware embedding composition scheme. Besides, we develop an innovative twin-attention network that alleviates the redundancy of the traditional multi-head self-attention while retaining full capacity for capturing long- and short-term (i.e., global and local) item dependencies. Comprehensive experiments demonstrate that LSAN significantly advances the accuracy and memory efficiency of existing sequential recommenders.

cs.IR↗

arXiv2021

Spin excitations in metallic kagome lattice FeSn and CoSn

In two-dimensional (2D) metallic kagome lattice materials, destructive interference of electronic hopping pathways around the kagome bracket can produce nearly localized electrons, and thus electronic bands that are flat in momentum space. When ferromagnetic order breaks the degeneracy of the electronic bands and splits them into the spin-up majority and spin-down minority electronic bands, quasiparticle excitations between the spin-up and spin-down flat bands should form a narrow localized spin-excitation Stoner continuum coexisting with well-defined spin waves in the long wavelengths. Here we report inelastic neutron scattering studies of spin excitations in 2D metallic Kagome lattice antiferromagnetic FeSn and paramagnetic CoSn, where angle resolved photoemission spectroscopy experiments found spin-polarized and nonpolarized flat bands, respectively, below the Fermi level. Although our initial measurements on FeSn indeed reveal well-defined spin waves extending well above 140 meV coexisting with a flat excitation at 170 meV, subsequent experiments on CoSn indicate that the flat mode actually arises mostly from hydrocarbon scattering of the CYTOP-M commonly used to glue the samples to aluminum holder. Therefore, our results established the evolution of spin excitations in FeSn and CoSn, and identified an anomalous flat mode that has been overlooked by the neutron scattering community for the past 20 years.

cond-mat.str-el↗

arXiv2021

GAG: Global Attributed Graph Neural Network for Streaming Session-based Recommendation

Streaming session-based recommendation (SSR) is a challenging task that requires the recommender system to do the session-based recommendation (SR) in the streaming scenario. In the real-world applications of e-commerce and social media, a sequence of user-item interactions generated within a certain period are grouped as a session, and these sessions consecutively arrive in the form of streams. Most of the recent SR research has focused on the static setting where the training data is first acquired and then used to train a session-based recommender model. They need several epochs of training over the whole dataset, which is infeasible in the streaming setting. Besides, they can hardly well capture long-term user interests because of the neglect or the simple usage of the user information. Although some streaming recommendation strategies have been proposed recently, they are designed for streams of individual interactions rather than streams of sessions. In this paper, we propose a Global Attributed Graph (GAG) neural network model with a Wasserstein reservoir for the SSR problem. On one hand, when a new session arrives, a session graph with a global attribute is constructed based on the current session and its associate user. Thus, the GAG can take both the global attribute and the current session into consideration to learn more comprehensive representations of the session and the user, yielding a better performance in the recommendation. On the other hand, for the adaptation to the streaming session scenario, a Wasserstein reservoir is proposed to help preserve a representative sketch of the historical data. Extensive experiments on two real-world datasets have been conducted to verify the superiority of the GAG model compared with the state-of-the-art methods.

cs.IR↗

arXiv2021

Exploiting Positional Information for Session-based Recommendation

For present e-commerce platforms, session-based recommender systems are developed to predict users' preference for next-item recommendation. Although a session can usually reflect a user's current preference, a local shift of the user's intention within the session may still exist. Specifically, the interactions that take place in the early positions within a session generally indicate the user's initial intention, while later interactions are more likely to represent the latest intention. Such positional information has been rarely considered in existing methods, which restricts their ability to capture the significance of interactions at different positions. To thoroughly exploit the positional information within a session, a theoretical framework is developed in this paper to provide an in-depth analysis of the positional information. We formally define the properties of forward-awareness and backward-awareness to evaluate the ability of positional encoding schemes in capturing the initial and the latest intention. According to our analysis, existing positional encoding schemes are generally forward-aware only, which can hardly represent the dynamics of the intention in a session. To enhance the positional encoding scheme for the session-based recommendation, a dual positional encoding (DPE) is proposed to account for both forward-awareness and backward-awareness. Based on DPE, we propose a novel Positional Recommender (PosRec) model with a well-designed Position-aware Gated Graph Neural Network module to fully exploit the positional information for session-based recommendation tasks. Extensive experiments are conducted on two e-commerce benchmark datasets, Yoochoose and Diginetica and the experimental results show the superiority of the PosRec by comparing it with the state-of-the-art session-based recommender models.

cs.IR↗

arXiv2021

Attribute-aware Explainable Complementary Clothing Recommendation

Modelling mix-and-match relationships among fashion items has become increasingly demanding yet challenging for modern E-commerce recommender systems. When performing clothes matching, most existing approaches leverage the latent visual features extracted from fashion item images for compatibility modelling, which lacks explainability of generated matching results and can hardly convince users of the recommendations. Though recent methods start to incorporate pre-defined attribute information (e.g., colour, style, length, etc.) for learning item representations and improving the model interpretability, their utilisation of attribute information is still mainly reserved for enhancing the learned item representations and generating explanations via post-processing. As a result, this creates a severe bottleneck when we are trying to advance the recommendation accuracy and generating fine-grained explanations since the explicit attributes have only loose connections to the actual recommendation process. This work aims to tackle the explainability challenge in fashion recommendation tasks by proposing a novel Attribute-aware Fashion Recommender (AFRec). Specifically, AFRec recommender assesses the outfit compatibility by explicitly leveraging the extracted attribute-level representations from each item's visual feature. The attributes serve as the bridge between two fashion items, where we quantify the affinity of a pair of items through the learned compatibility between their attributes. Extensive experiments have demonstrated that, by making full use of the explicit attributes in the recommendation process, AFRec is able to achieve state-of-the-art recommendation accuracy and generate intuitive explanations at the same time.

cs.IR↗

arXiv2021

Optimal sampling for design-based estimators of regression models

Two-phase designs measure variables of interest on a subcohort where the outcome and covariates are readily available or cheap to collect on all individuals in the cohort. Given limited resource availability, it is of interest to find an optimal design that includes more informative individuals in the final sample. We explore the optimal designs and efficiencies for analysis by design-based estimators. Generalized raking is an efficient design-based estimator that improves on the inverse-probability weighted (IPW) estimator by adjusting weights based on the auxiliary information. We derive a closed-form solution of the optimal design for estimating regression coefficients from generalized raking estimators. We compare it with the optimal design for analysis via the IPW estimator and other two-phase designs in measurement-error settings. We consider general two-phase designs where the outcome variable and variables of interest can be continuous or discrete. Our results show that the optimal designs for analysis by the two design-based estimators can be very different. The optimal design for IPW estimation is optimal for analysis via the IPW estimator and typically gives near-optimal efficiency for generalized raking, though we show there is potential improvement in some settings.

stat.ME↗