arXiv subjects

Fei Wang

Publications and source records attributed to Fei Wang.

At least 613 records · Page 34Linked to original sources

arXiv2020

A(DP)$^2$SGD: Asynchronous Decentralized Parallel Stochastic Gradient Descent with Differential Privacy

As deep learning models are usually massive and complex, distributed learning is essential for increasing training efficiency. Moreover, in many real-world application scenarios like healthcare, distributed learning can also keep the data local and protect privacy. A popular distributed learning strategy is federated learning, where there is a central server storing the global model and a set of local computing nodes updating the model parameters with their corresponding data. The updated model parameters will be processed and transmitted to the central server, which leads to heavy communication costs. Recently, asynchronous decentralized distributed learning has been proposed and demonstrated to be a more efficient and practical strategy where there is no central server, so that each computing node only communicates with its neighbors. Although no raw data will be transmitted across different local nodes, there is still a risk of information leak during the communication process for malicious participants to make attacks. In this paper, we present a differentially private version of asynchronous decentralized parallel SGD (ADPSGD) framework, or A(DP)$^2$SGD for short, which maintains communication efficiency of ADPSGD and prevents the inference from malicious participants. Specifically, R{é}nyi differential privacy is used to provide tighter privacy analysis for our composite Gaussian mechanisms while the convergence rate is consistent with the non-private version. Theoretical analysis shows A(DP)$^2$SGD also converges at the optimal $\mathcal{O}(1/\sqrt{T})$ rate as SGD. Empirically, A(DP)$^2$SGD achieves comparable model accuracy as the differentially private version of Synchronous SGD (SSGD) but runs much faster than SSGD in heterogeneous computing environments.

cs.LG↗

arXiv2020

Federated Learning for Healthcare Informatics

With the rapid development of computer software and hardware technologies, more and more healthcare data are becoming readily available from clinical institutions, patients, insurance companies and pharmaceutical industries, among others. This access provides an unprecedented opportunity for data science technologies to derive data-driven insights and improve the quality of care delivery. Healthcare data, however, are usually fragmented and private making it difficult to generate robust results across populations. For example, different hospitals own the electronic health records (EHR) of different patient populations and these records are difficult to share across hospitals because of their sensitive nature. This creates a big barrier for developing effective analytical approaches that are generalizable, which need diverse, "big data". Federated learning, a mechanism of training a shared global model with a central server while keeping all the sensitive data in local institutions where the data belong, provides great promise to connect the fragmented healthcare data sources with privacy-preservation. The goal of this survey is to provide a review for federated learning technologies, particularly within the biomedical space. In particular, we summarize the general solutions to the statistical challenges, system challenges and privacy issues in federated learning, and point out the implications and potentials in healthcare.

cs.LG↗

arXiv2020

Type-II neutrino seesaw mechanism extension of NMSSM from SUSY breaking mechanisms

We propose to accommodate economically the type-II neutrino seesaw mechanism in (G)NMSSM from GMSB and AMSB, respectively. The heavy triplets within neutrino seesaw mechanism are identified to be the messengers. Therefore, the $μ$-problem, the neutrino mass generation, LFV as well the soft SUSY breaking parameters can be economically combined in a non-trivial way. General features of such extensions are discussed. The type-II neutrino seesaw-specific interactions can give additional Yukawa deflection contributions to the soft SUSY breaking parameters of NMSSM, which are indispensable to realize successful EWSB and accommodate the 125 GeV Higgs. Relevant numerical results, including the constraints of dark matter and possible LFV processes $l_i\rightarrow l_j γ$ etc, are also given. We find that our economical type-II neutrino seesaw mechanism extension of NMSSM from AMSB or GMSB can lead to realistic low energy NMSSM spectrum, both admitting the 125 GeV Higgs as the lightest CP-even scalar. The possibility of the 125 GeV Higgs being the next-to-lightest CP-even scalar in GMSB-type scenario is ruled out by the constraints from EWSB, collider and precision measurements. The possibility of the 125 GeV Higgs being the next-to-lightest CP-even scalar in AMSB-type scenario is ruled out by dark matter direct detection experiments. Possible constraints from LFV processes $l_i\rightarrow l_j γ$ can give an upper bound for the messenger scale.

hep-ph↗

arXiv2020

Dynamic Knowledge Distillation for Black-box Hypothesis Transfer Learning

In real world applications like healthcare, it is usually difficult to build a machine learning prediction model that works universally well across different institutions. At the same time, the available model is often proprietary, i.e., neither the model parameter nor the data set used for model training is accessible. In consequence, leveraging the knowledge hidden in the available model (aka. the hypothesis) and adapting it to a local data set becomes extremely challenging. Motivated by this situation, in this paper we aim to address such a specific case within the hypothesis transfer learning framework, in which 1) the source hypothesis is a black-box model and 2) the source domain data is unavailable. In particular, we introduce a novel algorithm called dynamic knowledge distillation for hypothesis transfer learning (dkdHTL). In this method, we use knowledge distillation with instance-wise weighting mechanism to adaptively transfer the "dark" knowledge from the source hypothesis to the target domain.The weighting coefficients of the distillation loss and the standard loss are determined by the consistency between the predicted probability of the source hypothesis and the target ground-truth label.Empirical results on both transfer learning benchmark datasets and a healthcare dataset demonstrate the effectiveness of our method.

cs.LG↗

arXiv2020

Adversarial Infidelity Learning for Model Interpretation

Model interpretation is essential in data mining and knowledge discovery. It can help understand the intrinsic model working mechanism and check if the model has undesired characteristics. A popular way of performing model interpretation is Instance-wise Feature Selection (IFS), which provides an importance score of each feature representing the data samples to explain how the model generates the specific output. In this paper, we propose a Model-agnostic Effective Efficient Direct (MEED) IFS framework for model interpretation, mitigating concerns about sanity, combinatorial shortcuts, model identifiability, and information transmission. Also, we focus on the following setting: using selected features to directly predict the output of the given model, which serves as a primary evaluation metric for model-interpretation methods. Apart from the features, we involve the output of the given model as an additional input to learn an explainer based on more accurate information. To learn the explainer, besides fidelity, we propose an Adversarial Infidelity Learning (AIL) mechanism to boost the explanation learning by screening relatively unimportant features. Through theoretical and experimental analysis, we show that our AIL mechanism can help learn the desired conditional distribution between selected features and targets. Moreover, we extend our framework by integrating efficient interpretation methods as proper priors to provide a warm start. Comprehensive empirical evaluation results are provided by quantitative metrics and human evaluation to demonstrate the effectiveness and superiority of our proposed method. Our code is publicly available online at https://github.com/langlrsw/MEED.

stat.ML↗

arXiv2020

Explaining The XENON1T Excess With Light Goldstini Dark Matter

In the scenario with a multiplicity of sectors which independently break supersymmetry, multiplicity of goldstini are predicted. We propose a new interpretation of the electron recoil excess at 2-7 keV observed in the XENON1T experiment with very long-lived goldstini DM elastically scattering off the electrons. The goldstini DM can be boosted by the late-decay of the other nearly degenerate (long-lived) goldstini DM, with their tiny mass difference being converted into kinetic energy of the lighter goldstini DM and neutrinos. We show that viable parameter space can be found which can explain the excess of electron recoil events around 2-3 keV recently reported by the XENON1T experiment.

hep-ph↗

arXiv2020

Visualizing Deep Graph Generative Models for Drug Discovery

Drug discovery aims at designing novel molecules with specific desired properties for clinical trials. Over past decades, drug discovery and development have been a costly and time consuming process. Driven by big chemical data and AI, deep generative models show great potential to accelerate the drug discovery process. Existing works investigate different deep generative frameworks for molecular generation, however, less attention has been paid to the visualization tools to quickly demo and evaluate model's results. Here, we propose a visualization framework which provides interactive visualization tools to visualize molecules generated during the encoding and decoding process of deep graph generative models, and provide real time molecular optimization functionalities. Our work tries to empower black box AI driven drug discovery models with some visual interpretabilities.

cs.LG↗

arXiv2020

Coreference Resolution as Query-based Span Prediction

In this paper, we present an accurate and extensible approach for the coreference resolution task. We formulate the problem as a span prediction task, like in machine reading comprehension (MRC): A query is generated for each candidate mention using its surrounding context, and a span prediction module is employed to extract the text spans of the coreferences within the document using the generated query. This formulation comes with the following key advantages: (1) The span prediction strategy provides the flexibility of retrieving mentions left out at the mention proposal stage; (2) In the MRC framework, encoding the mention and its context explicitly in a query makes it possible to have a deep and thorough examination of cues embedded in the context of coreferent mentions; and (3) A plethora of existing MRC datasets can be used for data augmentation to improve the model's generalization capability. Experiments demonstrate significant performance boost over previous models, with 87.5 (+2.5) F1 score on the GAP benchmark and 83.1 (+3.5) F1 score on the CoNLL-2012 benchmark.

cs.CL↗

arXiv2020

Explicit-Blurred Memory Network for Analyzing Patient Electronic Health Records

In recent years, we have witnessed an increased interest in temporal modeling of patient records from large scale Electronic Health Records (EHR). While simpler RNN models have been used for such problems, memory networks, which in other domains were found to generalize well, are underutilized. Traditional memory networks involve diffused and non-linear operations where influence of past events on outputs are not readily quantifiable. We posit that this lack of interpretability makes such networks not applicable for EHR analysis. While networks with explicit memory have been proposed recently, the discontinuities imposed by the discrete operations make such networks harder to train and require more supervision. The problem is further exacerbated in the limited data setting of EHR studies. In this paper, we propose a novel memory architecture that is more interpretable than traditional memory networks while being easier to train than explicit memory banks. Inspired by well-known models of human cognition, we propose partitioning the external memory space into (a) a primary explicit memory block to store exact replicas of recent events to support interpretations, followed by (b) a secondary blurred memory block that accumulates salient aspects of past events dropped from the explicit block as higher level abstractions and allow training with less supervision by stabilize the gradients. We apply the model for 3 learning problems on ICU records from the MIMIC III database spanning millions of data points. Our model performs comparably to the state-of the art while also, crucially, enabling ready interpretation of the results.

cs.LG↗

arXiv2020

Compressible Navier-Stokes equations with heterogeneous pressure laws

This paper concerns the existence of global weak solutions à la Leray for compressible Navier-Stokes equations with a pressure law that depends on the density and on time and space variables $t$ and $x$. The assumptions on the pressure contain only locally Lipschitz assumption with respect to the density variable and some hypothesis with respect to the extra time and space variables. It may be seen as a first step to consider heat-conducting Navier-Stokes equations with physical laws such as the truncated virial assumption. The paper focuses on the construction of approximate solutions through a new regularized and fixed point procedure and on the weak stability process taking advantage of the new method introduced by the two first authors with a careful study of an appropriate regularized quantity linked to the pressure.

math.AP↗

arXiv2020

Non-Markovian trajectories involving future in the semi-classical path integral expression

Semiclassical path integral expression for a quantum system coupled to a harmonic bath is derived based on the stationary phase condition. It is discovered that the system path is non-Markovian. Most strikingly, the system path not only couples to its past (as in the Langevin equation), but also to its future, i.e. the equation of motion for the system is an integro-differential equation that involves all times. Numerical tests are performed to confirm that the future-involved term is indeed necessary. Because of the future-non-Markovian nature of the equation, the numerical solution cannot be obtained by iterative methods. Instead, root search algorithms must be employed.

quant-ph↗

arXiv2020

Neural Dynamics on Complex Networks

Learning continuous-time dynamics on complex networks is crucial for understanding, predicting and controlling complex systems in science and engineering. However, this task is very challenging due to the combinatorial complexities in the structures of high dimensional systems, their elusive continuous-time nonlinear dynamics, and their structural-dynamic dependencies. To address these challenges, we propose to combine Ordinary Differential Equation Systems (ODEs) and Graph Neural Networks (GNNs) to learn continuous-time dynamics on complex networks in a data-driven manner. We model differential equation systems by GNNs. Instead of mapping through a discrete number of neural layers in the forward process, we integrate GNN layers over continuous time numerically, leading to capturing continuous-time dynamics on graphs. Our model can be interpreted as a Continuous-time GNN model or a Graph Neural ODEs model. Our model can be utilized for continuous-time network dynamics prediction, structured sequence prediction (a regularly-sampled case), and node semi-supervised classification tasks (a one-snapshot case) in a unified framework. We validate our model by extensive experiments in the above three scenarios. The promising experimental results demonstrate our model's capability of jointly capturing the structure and dynamics of complex systems in a unified framework.

cs.SI↗

arXiv2020

MoFlow: An Invertible Flow Model for Generating Molecular Graphs

Generating molecular graphs with desired chemical properties driven by deep graph generative models provides a very promising way to accelerate drug discovery process. Such graph generative models usually consist of two steps: learning latent representations and generation of molecular graphs. However, to generate novel and chemically-valid molecular graphs from latent representations is very challenging because of the chemical constraints and combinatorial complexity of molecular graphs. In this paper, we propose MoFlow, a flow-based graph generative model to learn invertible mappings between molecular graphs and their latent representations. To generate molecular graphs, our MoFlow first generates bonds (edges) through a Glow based model, then generates atoms (nodes) given bonds by a novel graph conditional flow, and finally assembles them into a chemically valid molecular graph with a posthoc validity correction. Our MoFlow has merits including exact and tractable likelihood training, efficient one-pass embedding and generation, chemical validity guarantees, 100\% reconstruction of training data, and good generalization ability. We validate our model by four tasks: molecular graph generation and reconstruction, visualization of the continuous latent space, property optimization, and constrained property optimization. Our MoFlow achieves state-of-the-art performance, which implies its potential efficiency and effectiveness to explore large chemical space for drug discovery.

stat.ML↗

arXiv2020

Learning 3D-3D Correspondences for One-shot Partial-to-partial Registration

While 3D-3D registration is traditionally tacked by optimization-based methods, recent work has shown that learning-based techniques could achieve faster and more robust results. In this context, however, only PRNet can handle the partial-to-partial registration scenario. Unfortunately, this is achieved at the cost of relying on an iterative procedure, with a complex network architecture. Here, we show that learning-based partial-to-partial registration can be achieved in a one-shot manner, jointly reducing network complexity and increasing registration accuracy. To this end, we propose an Optimal Transport layer able to account for occluded points thanks to the use of outlier bins. The resulting OPRNet framework outperforms the state of the art on standard benchmarks, demonstrating better robustness and generalization ability than existing techniques.

cs.CV↗

arXiv2020

gPAV-Based Unconditionally Energy-Stable Schemes for the Cahn-Hilliard Equation: Stability and Error Analysis

We present several first-order and second-order numerical schemes for the Cahn-Hilliard equation with discrete unconditional energy stability. These schemes stem from the generalized Positive Auxiliary Variable (gPAV) idea, and require only the solution of linear algebraic systems with a constant coefficient matrix. More importantly, the computational complexity (operation count per time step) of these schemes is approximately a half of those of the gPAV and the scalar auxiliary variable (SAV) methods in previous works. We investigate the stability properties of the proposed schemes to establish stability bounds for the field function and the auxiliary variable, and also provide their error analyses. Numerical experiments are presented to verify the theoretical analyses and also demonstrate the stability of the schemes at large time step sizes.

math.NA↗

arXiv2020

General-Purpose User Embeddings based on Mobile App Usage

In this paper, we report our recent practice at Tencent for user modeling based on mobile app usage. User behaviors on mobile app usage, including retention, installation, and uninstallation, can be a good indicator for both long-term and short-term interests of users. For example, if a user installs Snapseed recently, she might have a growing interest in photographing. Such information is valuable for numerous downstream applications, including advertising, recommendations, etc. Traditionally, user modeling from mobile app usage heavily relies on handcrafted feature engineering, which requires onerous human work for different downstream applications, and could be sub-optimal without domain experts. However, automatic user modeling based on mobile app usage faces unique challenges, including (1) retention, installation, and uninstallation are heterogeneous but need to be modeled collectively, (2) user behaviors are distributed unevenly over time, and (3) many long-tailed apps suffer from serious sparsity. In this paper, we present a tailored AutoEncoder-coupled Transformer Network (AETN), by which we overcome these challenges and achieve the goals of reducing manual efforts and boosting performance. We have deployed the model at Tencent, and both online/offline experiments from multiple domains of downstream applications have demonstrated the effectiveness of the output user embeddings.

cs.LG↗

arXiv2020

Multi-Scale Boosted Dehazing Network with Dense Feature Fusion

In this paper, we propose a Multi-Scale Boosted Dehazing Network with Dense Feature Fusion based on the U-Net architecture. The proposed method is designed based on two principles, boosting and error feedback, and we show that they are suitable for the dehazing problem. By incorporating the Strengthen-Operate-Subtract boosting strategy in the decoder of the proposed model, we develop a simple yet effective boosted decoder to progressively restore the haze-free image. To address the issue of preserving spatial information in the U-Net architecture, we design a dense feature fusion module using the back-projection feedback scheme. We show that the dense feature fusion module can simultaneously remedy the missing spatial information from high-resolution features and exploit the non-adjacent features. Extensive evaluations demonstrate that the proposed model performs favorably against the state-of-the-art approaches on the benchmark datasets as well as real-world hazy images.

cs.CV↗

arXiv2020

A Real-Time Cross-modality Correlation Filtering Method for Referring Expression Comprehension

Referring expression comprehension aims to localize the object instance described by a natural language expression. Current referring expression methods have achieved good performance. However, none of them is able to achieve real-time inference without accuracy drop. The reason for the relatively slow inference speed is that these methods artificially split the referring expression comprehension into two sequential stages including proposal generation and proposal ranking. It does not exactly conform to the habit of human cognition. To this end, we propose a novel Realtime Cross-modality Correlation Filtering method (RCCF). RCCF reformulates the referring expression comprehension as a correlation filtering process. The expression is first mapped from the language domain to the visual domain and then treated as a template (kernel) to perform correlation filtering on the image feature map. The peak value in the correlation heatmap indicates the center points of the target box. In addition, RCCF also regresses a 2-D object size and 2-D offset. The center point coordinates, object size and center point offset together to form the target bounding box. Our method runs at 40 FPS while achieving leading performance in RefClef, RefCOCO, RefCOCO+ and RefCOCOg benchmarks. In the challenging RefClef dataset, our methods almost double the state-of-the-art performance (34.70% increased to 63.79%). We hope this work can arouse more attention and studies to the new cross-modality correlation filtering framework as well as the one-stage framework for referring expression comprehension.

cs.CV↗