arXiv subjects

Yan Liu

Publications and source records attributed to Yan Liu.

At least 343 records · Page 19Linked to original sources

arXiv2021

Sparse principal component analysis for high-dimensional stationary time series

We consider the sparse principal component analysis for high-dimensional stationary processes. The standard principal component analysis performs poorly when the dimension of the process is large. We establish the oracle inequalities for penalized principal component estimators for the processes including heavy-tailed time series. The rate of convergence of the estimators is established. We also elucidate the theoretical rate for choosing the tuning parameter in penalized estimators. The performance of the sparse principal component analysis is demonstrated by numerical simulations. The utility of the sparse principal component analysis for time series data is exemplified by the application to average temperature data.

math.ST↗

arXiv2021

Non-autoregressive Transformer with Unified Bidirectional Decoder for Automatic Speech Recognition

Non-autoregressive (NAR) transformer models have been studied intensively in automatic speech recognition (ASR), and a substantial part of NAR transformer models is to use the casual mask to limit token dependencies. However, the casual mask is designed for the left-to-right decoding process of the non-parallel autoregressive (AR) transformer, which is inappropriate for the parallel NAR transformer since it ignores the right-to-left contexts. Some models are proposed to utilize right-to-left contexts with an extra decoder, but these methods increase the model complexity. To tackle the above problems, we propose a new non-autoregressive transformer with a unified bidirectional decoder (NAT-UBD), which can simultaneously utilize left-to-right and right-to-left contexts. However, direct use of bidirectional contexts will cause information leakage, which means the decoder output can be affected by the character information from the input of the same position. To avoid information leakage, we propose a novel attention mask and modify vanilla queries, keys, and values matrices for NAT-UBD. Experimental results verify that NAT-UBD can achieve character error rates (CERs) of 5.0%/5.5% on the Aishell1 dev/test sets, outperforming all previous NAR transformer models. Moreover, NAT-UBD can run 49.8x faster than the AR transformer baseline when decoding in a single step.

cs.CL↗

arXiv2021

GANSER: A Self-supervised Data Augmentation Framework for EEG-based Emotion Recognition

The data scarcity problem in Electroencephalography (EEG) based affective computing results into difficulty in building an effective model with high accuracy and stability using machine learning algorithms especially deep learning models. Data augmentation has recently achieved considerable performance improvement for deep learning models: increased accuracy, stability, and reduced over-fitting. In this paper, we propose a novel data augmentation framework, namely Generative Adversarial Network-based Self-supervised Data Augmentation (GANSER). As the first to combine adversarial training with self-supervised learning for EEG-based emotion recognition, the proposed framework can generate high-quality and high-diversity simulated EEG samples. In particular, we utilize adversarial training to learn an EEG generator and force the generated EEG signals to approximate the distribution of real samples, ensuring the quality of augmented samples. A transformation function is employed to mask parts of EEG signals and force the generator to synthesize potential EEG signals based on the remaining parts, to produce a wide variety of samples. The masking possibility during transformation is introduced as prior knowledge to guide to extract distinguishable features for simulated EEG signals and generalize the classifier to the augmented sample space. Finally, extensive experiments demonstrate our proposed method can help emotion recognition for performance gain and achieve state-of-the-art results.

cs.LG↗

arXiv2021

MeDiaQA: A Question Answering Dataset on Medical Dialogues

In this paper, we introduce MeDiaQA, a novel question answering(QA) dataset, which constructed on real online Medical Dialogues. It contains 22k multiple-choice questions annotated by human for over 11k dialogues with 120k utterances between patients and doctors, covering 150 specialties of diseases, which are collected from haodf.com and dxy.com. MeDiaQA is the first QA dataset where reasoning over medical dialogues, especially their quantitative contents. The dataset has the potential to test the computing, reasoning and understanding ability of models across multi-turn dialogues, which is challenging compared with the existing datasets. To address the challenges, we design MeDia-BERT, and it achieves 64.3% accuracy, while human performance of 93% accuracy, which indicates that there still remains a large room for improvement.

cs.CL↗

arXiv2021

Statistical Inference for Local Granger Causality

Granger causality has been employed to investigate causality relations between components of stationary multiple time series. We generalize this concept by developing statistical inference for local Granger causality for multivariate locally stationary processes. Our proposed local Granger causality approach captures time-evolving causality relationships in nonstationary processes. The proposed local Granger causality is well represented in the frequency domain and estimated based on the parametric time-varying spectral density matrix using the local Whittle likelihood. Under regularity conditions, we demonstrate that the estimators converge to multivariate normal in distribution. Additionally, the test statistic for the local Granger causality is shown to be asymptotically distributed as a quadratic form of a multivariate normal distribution. The finite sample performance is confirmed with several simulation studies for multivariate time-varying autoregressive models. For practical demonstration, the proposed local Granger causality method uncovered new functional connectivity relationships between channels in brain signals. Moreover, the method was able to identify structural changes in financial data.

stat.ME↗

arXiv2021

Holographic p-wave superconductivity from higher derivative theory

We construct a holographic SU(2) p-wave superconductor model with Weyl corrections. The high derivative (HD) terms do not seem to spoil the generation of the p-wave superconducting phase. We mainly study the properties of AC conductivity, which is absent in holographic SU(2) p-wave superconductor with Weyl corrections. The conductivities in superconducting phase exhibit obvious anisotropic behaviors. Along $y$ direction, the conductivity $σ_{yy}$ is similar to that of holographic s-wave superconductor. The superconducting energy gap exhibits a wide extension. For the conductivity $σ_{xx}$ along $x$ direction, the behaviors of the real part in the normal state are closely similar to that of $σ_{yy}$. However, the anisotropy of the conductivity obviously shows up in the superconducting phase. A Drude-like peak at low frequency emerges in $Reσ_{xx}$ once the system enters into the superconducting phase, regardless of the behaviors in normal state.

hep-th↗

arXiv2021

Phoneme-aware and Channel-wise Attentive Learning for Text DependentSpeaker Verification

This paper proposes a multi-task learning network with phoneme-aware and channel-wise attentive learning strategies for text-dependent Speaker Verification (SV). In the proposed structure, the frame-level multi-task learning along with the segment-level adversarial learning is adopted for speaker embedding extraction. The phoneme-aware attentive pooling is exploited on frame-level features in the main network for speaker classifier, with the corresponding posterior probability for the phoneme distribution in the auxiliary subnet. Further, the introduction of Squeeze and Excitation (SE-block) performs dynamic channel-wise feature recalibration, which improves the representational ability. The proposed method exploits speaker idiosyncrasies associated with pass-phrases, and is further improved by the phoneme-aware attentive pooling and SE-block from temporal and channel-wise aspects, respectively. The experiments conducted on RSR2015 Part 1 database confirm that the proposed system achieves outstanding results for textdependent SV.

cs.SD↗

arXiv2021

Additive Phoneme-aware Margin Softmax Loss for Language Recognition

This paper proposes an additive phoneme-aware margin softmax (APM-Softmax) loss to train the multi-task learning network with phonetic information for language recognition. In additive margin softmax (AM-Softmax) loss, the margin is set as a constant during the entire training for all training samples, and that is a suboptimal method since the recognition difficulty varies in training samples. In additive angular margin softmax (AAM-Softmax) loss, the additional angular margin is set as a costant as well. In this paper, we propose an APM-Softmax loss for language recognition with phoneitc multi-task learning, in which the additive phoneme-aware margin is automatically tuned for different training samples. More specifically, the margin of language recognition is adjusted according to the results of phoneme recognition. Experiments are reported on Oriental Language Recognition (OLR) datasets, and the proposed method improves AM-Softmax loss and AAM-Softmax loss in different language recognition testing conditions.

cs.SD↗

arXiv2021

MetaDetector: Meta Event Knowledge Transfer for Fake News Detection

The blooming of fake news on social networks has devastating impacts on society, economy, and public security. Although numerous studies are conducted for the automatic detection of fake news, the majority tend to utilize deep neural networks to learn event-specific features for superior detection performance on specific datasets. However, the trained models heavily rely on the training datasets and are infeasible to apply to upcoming events due to the discrepancy between event distributions. Inspired by domain adaptation theories, we propose an end-to-end adversarial adaptation network, dubbed as MetaDetector, to transfer meta knowledge (event-shared features) between different events. Specifically, MetaDetector pushes the feature extractor and event discriminator to eliminate event-specific features and preserve required event-shared features by adversarial training. Furthermore, the pseudo-event discriminator is utilized to evaluate the importance of historical event posts to obtain partial shared features that are discriminative for detecting fake news. Under the coordinated optimization among the four submodules, MetaDetector accurately transfers the meta-knowledge of historical events to the upcoming event for fact checking. We conduct extensive experiments on two large-scale datasets collected from Weibo and Twitter. The experimental results demonstrate that MetaDetector outperforms the state-of-the-art methods, especially when the distribution shift between events is significant. Furthermore, we find that MetaDetector is able to learn the event-shared features, and alleviate the negative transfer caused by the large distribution shift between events.

cs.SI↗

arXiv2021

Cross-Node Federated Graph Neural Network for Spatio-Temporal Data Modeling

Vast amount of data generated from networks of sensors, wearables, and the Internet of Things (IoT) devices underscores the need for advanced modeling techniques that leverage the spatio-temporal structure of decentralized data due to the need for edge computation and licensing (data access) issues. While federated learning (FL) has emerged as a framework for model training without requiring direct data sharing and exchange, effectively modeling the complex spatio-temporal dependencies to improve forecasting capabilities still remains an open problem. On the other hand, state-of-the-art spatio-temporal forecasting models assume unfettered access to the data, neglecting constraints on data sharing. To bridge this gap, we propose a federated spatio-temporal model -- Cross-Node Federated Graph Neural Network (CNFGNN) -- which explicitly encodes the underlying graph structure using graph neural network (GNN)-based architecture under the constraint of cross-node federated learning, which requires that data in a network of nodes is generated locally on each node and remains decentralized. CNFGNN operates by disentangling the temporal dynamics modeling on devices and spatial dynamics on the server, utilizing alternating optimization to reduce the communication cost, facilitating computations on the edge devices. Experiments on the traffic flow forecasting task show that CNFGNN achieves the best forecasting performance in both transductive and inductive learning settings with no extra computation cost on edge devices, while incurring modest communication cost.

cs.LG↗

arXiv2021

Physics-aware Spatiotemporal Modules with Auxiliary Tasks for Meta-Learning

Modeling the dynamics of real-world physical systems is critical for spatiotemporal prediction tasks, but challenging when data is limited. The scarcity of real-world data and the difficulty in reproducing the data distribution hinder directly applying meta-learning techniques. Although the knowledge of governing partial differential equations (PDE) of data can be helpful for the fast adaptation to few observations, it is mostly infeasible to exactly find the equation for observations in real-world physical systems. In this work, we propose a framework, physics-aware meta-learning with auxiliary tasks, whose spatial modules incorporate PDE-independent knowledge and temporal modules utilize the generalized features from the spatial modules to be adapted to the limited data, respectively. The framework is inspired by a local conservation law expressed mathematically as a continuity equation and does not require the exact form of governing equation to model the spatiotemporal observations. The proposed method mitigates the need for a large number of real-world tasks for meta-learning by leveraging spatial information in simulated data to meta-initialize the spatial modules. We apply the proposed framework to both synthetic and real-world spatiotemporal prediction tasks and demonstrate its superior performance with limited observations.

cs.LG↗

arXiv2021

Variational Inference for Latent Space Models for Dynamic Networks

Latent space models are popular for analyzing dynamic network data. We propose a variational approach to estimate the model parameters as well as the latent positions of the nodes in the network. The variational approach is much faster than Markov chain Monte Carlo algorithms, and is able to handle large networks. Theoretical properties of the variational Bayes risk of the proposed procedure are provided. We apply the variational method and latent space model to simulated data as well as real data to demonstrate its performance.

stat.ME↗

arXiv2021

An improved holographic nodal line semimetal

We study an improved holographic model for the strongly coupled nodal line semimetal which satisfies the duality relation between the rank two tensor operators $\barψγ^{μν}ψ$ and $\barψγ^{μν}γ^5ψ$. We introduce a Chern-Simons term and a mass term in the bulk for a complex two form field which is dual to the above tensor operators and the duality relation is automatically satisfied from holography. We find that there exists a quantum phase transition from a topological nodal line semimetal phase to a trivial phase. In the topological phase, there exist multiple nodal lines in the fermionic spectrum which are topologically nontrivial. The bulk geometries are different from the previous model without the duality constraint, while the resulting properties are qualitatively similar to those in that model. This improved model provides a more natural ground to analyze transports or other properties of strongly coupled nodal line semimetals.

hep-th↗

arXiv2021

Identifying Coordinated Accounts on Social Media through Hidden Influence and Group Behaviours

Disinformation campaigns on social media, involving coordinated activities from malicious accounts towards manipulating public opinion, have become increasingly prevalent. Existing approaches to detect coordinated accounts either make very strict assumptions about coordinated behaviours, or require part of the malicious accounts in the coordinated group to be revealed in order to detect the rest. To address these drawbacks, we propose a generative model, AMDN-HAGE (Attentive Mixture Density Network with Hidden Account Group Estimation) which jointly models account activities and hidden group behaviours based on Temporal Point Processes (TPP) and Gaussian Mixture Model (GMM), to capture inherent characteristics of coordination which is, accounts that coordinate must strongly influence each other's activities, and collectively appear anomalous from normal accounts. To address the challenges of optimizing the proposed model, we provide a bilevel optimization algorithm with theoretical guarantee on convergence. We verified the effectiveness of the proposed method and training algorithm on real-world social network data collected from Twitter related to coordinated campaigns from Russia's Internet Research Agency targeting the 2016 U.S. Presidential Elections, and to identify coordinated campaigns related to the COVID-19 pandemic. Leveraging the learned model, we find that the average influence between coordinated account pairs is the highest.On COVID-19, we found coordinated group spreading anti-vaccination, anti-masks conspiracies that suggest the pandemic is a hoax and political scam.

cs.SI↗

arXiv2021

One Network to Solve Them All: A Sequential Multi-Task Joint Learning Network Framework for MR Imaging Pipeline

Magnetic resonance imaging (MRI) acquisition, reconstruction, and segmentation are usually processed independently in the conventional practice of MRI workflow. It is easy to notice that there are significant relevances among these tasks and this procedure artificially cuts off these potential connections, which may lead to losing clinically important information for the final diagnosis. To involve these potential relations for further performance improvement, a sequential multi-task joint learning network model is proposed to train a combined end-to-end pipeline in a differentiable way, aiming at exploring the mutual influence among those tasks simultaneously. Our design consists of three cascaded modules: 1) deep sampling pattern learning module optimizes the $k$-space sampling pattern with predetermined sampling rate; 2) deep reconstruction module is dedicated to reconstructing MR images from the undersampled data using the learned sampling pattern; 3) deep segmentation module encodes MR images reconstructed from the previous module to segment the interested tissues. The proposed model retrieves the latently interactive and cyclic relations among those tasks, from which each task will be mutually beneficial. The proposed framework is verified on MRB dataset, which achieves superior performance on other SOTA methods in terms of both reconstruction and segmentation.

eess.IV↗

arXiv2021

PolSIRD: Modeling Epidemic Spread under Intervention Policies

Epidemic spread in a population is traditionally modeled via compartmentalized models which represent the free evolution of disease in absence of any intervention policies. In addition, these models assume full observability of disease cases and do not account for under-reporting. We present a mathematical model, namely PolSIRD, which accounts for the under-reporting by introducing an observation mechanism. It also captures the effects of intervention policies on the disease spread parameters by leveraging intervention policy data along with the reported disease cases. Furthermore, we allow our recurrent model to learn the initial hidden state of all compartments end-to-end along with other parameters via gradient-based training. We apply our model to the spread of the recent global outbreak of COVID-19 in the United States, where our model outperforms the methods employed by the CDC in predicting the spread. We also provide counterfactual simulations from our model to analyze the effect of lifting the intervention policies prematurely and our model correctly predicts the second wave of the epidemic.

q-bio.PE↗

arXiv2021

An Examination of Fairness of AI Models for Deepfake Detection

Recent studies have demonstrated that deep learning models can discriminate based on protected classes like race and gender. In this work, we evaluate bias present in deepfake datasets and detection models across protected subgroups. Using facial datasets balanced by race and gender, we examine three popular deepfake detectors and find large disparities in predictive performances across races, with up to 10.7% difference in error rate between subgroups. A closer look reveals that the widely used FaceForensics++ dataset is overwhelmingly composed of Caucasian subjects, with the majority being female Caucasians. Our investigation of the racial distribution of deepfakes reveals that the methods used to create deepfakes as positive training signals tend to produce "irregular" faces - when a person's face is swapped onto another person of a different race or gender. This causes detectors to learn spurious correlations between the foreground faces and fakeness. Moreover, when detectors are trained with the Blended Image (BI) dataset from Face X-Rays, we find that those detectors develop systematic discrimination towards certain racial subgroups, primarily female Asians.

cs.CV↗

arXiv2021

Deterministic Assembly of Single Emitters in Sub-5 Nanometer Optical Cavity Formed by Gold Nanorod Dimers on Three-Dimensional DNA Origami

Controllable strong interactions between a nanocavity and a single emitter is important to manipulating optical emission in a nanophotonic systems but challenging to achieve. Here a three-dimensional DNA origami, named as DNA rack (DR) is proposed and demonstrated to deterministically and precisely assemble single emitters within ultra-small plasmonic nanocavities formed by closely coupled gold nanorods (AuNRs). The DR uniquely possesses a saddle shape with two tubular grooves that geometrically allows a snug fit and linearly align two AuNRs with a bending angle <10°. It also includes a spacer at the saddle point to maintain the gap between AuNRs as small as 2-3 nm, forming a nanocavity estimated to be 20 nm3 and an experimentally measured Q factor of 7.3. A DNA docking strand is designed at the spacer to position a single fluorescent emitter at nanometer accuracy within the cavity. Using Cy5 as a model emitter, a ~30-fold fluorescence enhancement and a significantly reduced emission lifetime (from 1.6 ns to 670 ps) were experimentally verified, confirming significant emitter-cavity interactions. This DR-templated assembly method is capable of fitting AuNRs of variable length-to-width aspect ratios to form anisotropic nanocavities and deterministically incorporating different single emitters, thus enabling flexible design of both cavity resonance and emission wavelengths to tailor light-matter interactions at nanometer scale.

physics.optics↗