arXiv subjects

Cheng Zhang

Publications and source records attributed to Cheng Zhang.

At least 433 records · Page 24Linked to original sources

arXiv2020

Data Manipulation: Towards Effective Instance Learning for Neural Dialogue Generation via Learning to Augment and Reweight

Current state-of-the-art neural dialogue models learn from human conversations following the data-driven paradigm. As such, a reliable training corpus is the crux of building a robust and well-behaved dialogue model. However, due to the open-ended nature of human conversations, the quality of user-generated training data varies greatly, and effective training samples are typically insufficient while noisy samples frequently appear. This impedes the learning of those data-driven neural dialogue models. Therefore, effective dialogue learning requires not only more reliable learning samples, but also fewer noisy samples. In this paper, we propose a data manipulation framework to proactively reshape the data distribution towards reliable samples by augmenting and highlighting effective learning samples as well as reducing the effect of inefficient samples simultaneously. In particular, the data manipulation model selectively augments the training samples and assigns an importance weight to each instance to reform the training data. Note that, the proposed data manipulation framework is fully data-driven and learnable. It not only manipulates training samples to optimize the dialogue generation model, but also learns to increase its manipulation skills through gradient descent with validation samples. Extensive experiments show that our framework can improve the dialogue generation performance with respect to various automatic evaluation metrics and human judgments.

cs.CL↗

arXiv2020

Non-bifurcating phylogenetic tree inference via the adaptive LASSO

Phylogenetic tree inference using deep DNA sequencing is reshaping our understanding of rapidly evolving systems, such as the within-host battle between viruses and the immune system. Densely sampled phylogenetic trees can contain special features, including "sampled ancestors" in which we sequence a genotype along with its direct descendants, and "polytomies" in which multiple descendants arise simultaneously. These features are apparent after identifying zero-length branches in the tree. However, current maximum-likelihood based approaches are not capable of revealing such zero-length branches. In this paper, we find these zero-length branches by introducing adaptive-LASSO-type regularization estimators to phylogenetics, deriving their properties, and showing regularization to be a practically useful approach for phylogenetics.

q-bio.PE↗

arXiv2020

Attention-based network for low-light image enhancement

The captured images under low light conditions often suffer insufficient brightness and notorious noise. Hence, low-light image enhancement is a key challenging task in computer vision. A variety of methods have been proposed for this task, but these methods often failed in an extreme low-light environment and amplified the underlying noise in the input image. To address such a difficult problem, this paper presents a novel attention-based neural network to generate high-quality enhanced low-light images from the raw sensor data. Specifically, we first employ attention strategy (i.e. channel attention and spatial attention modules) to suppress undesired chromatic aberration and noise. The channel attention module guides the network to refine redundant colour features. The spatial attention module focuses on denoising by taking advantage of the non-local correlation in the image. Furthermore, we propose a new pooling layer, called inverted shuffle layer, which adaptively selects useful information from previous features. Extensive experiments demonstrate the superiority of the proposed network in terms of suppressing the chromatic aberration and noise artifacts in enhancement, especially when the low-light image has severe noise.

eess.IV↗

arXiv2020

Learning from Easy to Complex: Adaptive Multi-curricula Learning for Neural Dialogue Generation

Current state-of-the-art neural dialogue systems are mainly data-driven and are trained on human-generated responses. However, due to the subjectivity and open-ended nature of human conversations, the complexity of training dialogues varies greatly. The noise and uneven complexity of query-response pairs impede the learning efficiency and effects of the neural dialogue generation models. What is more, so far, there are no unified dialogue complexity measurements, and the dialogue complexity embodies multiple aspects of attributes---specificity, repetitiveness, relevance, etc. Inspired by human behaviors of learning to converse, where children learn from easy dialogues to complex ones and dynamically adjust their learning progress, in this paper, we first analyze five dialogue attributes to measure the dialogue complexity in multiple perspectives on three publicly available corpora. Then, we propose an adaptive multi-curricula learning framework to schedule a committee of the organized curricula. The framework is established upon the reinforcement learning paradigm, which automatically chooses different curricula at the evolving learning process according to the learning status of the neural dialogue generation model. Extensive experiments conducted on five state-of-the-art models demonstrate its learning efficiency and effectiveness with respect to 13 automatic evaluation metrics and human judgments.

cs.CL↗

arXiv2020

The Collectivity of Heavy Mesons in Proton-Nucleus Collisions

Using a model based on the Color Glass Condensate framework and the dilute-dense factorization, we systematically study the azimuthal angular correlations between a heavy flavor meson and a light reference particle in proton-nucleus collisions. The obtained second harmonic coefficients (also known as the elliptic flows) for $J/ψ$ and $D^0$ agree with recent experimental data from the LHC. We also provide predictions for the elliptic flows of $Υ$ and $B$ meson, which can be measured in the near future at the LHC. This work can shed light on the physics origin of the collectivity phenomenon in the collisions of small systems.

hep-ph↗

arXiv2020

$L^p$ eigenfunction bounds for fractional Schrödinger operators on manifolds

This paper is dedicated to $L^p$ bounds on eigenfunctions of a Schödinger-type operator $(-Δ_g)^{α/2} +V$ on closed Riemannian manifolds for critically singular potentials $V$. The operator $(-Δ_g)^{α/2}$ is defined spectrally in terms of the eigenfunctions of $-Δ_g$. We obtain also quasimodes and spectral clusters estimates. As an application, we derive Strichartz estimates for the fractional wave equation $(\partial_t^2+(-Δ_g)^{α/2}+V)u=0$. The wave kernel techniques recently developed by Bourgain-Shao-Sogge-Yao and Shao-Yao play a key role in this paper. We construct a new reproducing operator with several local operators and some good error terms. Moreover, we shall prove that these local operators satisfy certain variable coefficient versions of the "uniform Sobolev estimates" by Kenig-Ruiz-Sogge. These enable us to handle the critically singular potentials $V$ and prove the quasimode estimates.

math.AP↗

arXiv2020

Eigenfunction equations of lattice KdV equations and connections to ABS lattice equations with a $δ$ term

We develop lattice eigenfunction equations of lattice KdV equation, which are equations obeyed by the auxiliary functions, or eigenfunctions, of the Lax pair of the lattice KdV equation. This leads to three-dimensionally consistent quad-equations that are closely related to lattice equations in the Adler-Bobenko-Suris (ABS) classification. In particular, we show how the H3($δ$), Q1($δ$) and Q3($δ$) equations in the ABS list arise from the lattice eigenfunction equations by providing a natural interpretation of the $δ$ term as interactions between the eigenfunctions. By construction, exact solution structures of these equations are obtained. The approach presented in this paper can be used as a systematic means to search for integrable lattice equations.

nlin.SI↗

arXiv2020

Assessing the Memory Ability of Recurrent Neural Networks

It is known that Recurrent Neural Networks (RNNs) can remember, in their hidden layers, part of the semantic information expressed by a sequence (e.g., a sentence) that is being processed. Different types of recurrent units have been designed to enable RNNs to remember information over longer time spans. However, the memory abilities of different recurrent units are still theoretically and empirically unclear, thus limiting the development of more effective and explainable RNNs. To tackle the problem, in this paper, we identify and analyze the internal and external factors that affect the memory ability of RNNs, and propose a Semantic Euclidean Space to represent the semantics expressed by a sequence. Based on the Semantic Euclidean Space, a series of evaluation indicators are defined to measure the memory abilities of different recurrent units and analyze their limitations. These evaluation indicators also provide a useful guidance to select suitable sequence lengths for different RNNs during training.

cs.LG↗

arXiv2020

Double-dot interferometer for quantum measurement of Majorana qubits and stabilizers

Motivated by the need of quantum measurement of Majorana qubits and surface-code stabilizers, we analyze the performance of a double-dot interferometer under the influence of environment noise. The double-dot setup design allows accounting for the full multiple tunneling process between the dots through the Majorana island, within a master equation approach. In the co-tunneling regime, which results in a Majorana-mediated effective coupling between the dots, the master equation approach allows us to obtain analytic solutions for the measurement currents. The measurement quality, characterized by figures of merit such as the visibility of measurement signals, is carried out in regard to the unusual decoherence effect rather than `which-path' dephasing. The results obtained in this work are expected to be useful for future experiments of Majorana qubit and stabilizer measurements.

quant-ph↗

arXiv2020

Adaptive Parameterization for Neural Dialogue Generation

Neural conversation systems generate responses based on the sequence-to-sequence (SEQ2SEQ) paradigm. Typically, the model is equipped with a single set of learned parameters to generate responses for given input contexts. When confronting diverse conversations, its adaptability is rather limited and the model is hence prone to generate generic responses. In this work, we propose an {\bf Ada}ptive {\bf N}eural {\bf D}ialogue generation model, \textsc{AdaND}, which manages various conversations with conversation-specific parameterization. For each conversation, the model generates parameters of the encoder-decoder by referring to the input context. In particular, we propose two adaptive parameterization mechanisms: a context-aware and a topic-aware parameterization mechanism. The context-aware parameterization directly generates the parameters by capturing local semantics of the given context. The topic-aware parameterization enables parameter sharing among conversations with similar topics by first inferring the latent topics of the given context and then generating the parameters with respect to the distributional topics. Extensive experiments conducted on a large-scale real-world conversational dataset show that our model achieves superior performance in terms of both quantitative metrics and human evaluations.

cs.CL↗

arXiv2019

Restriction of toral eigenfunctions to totally geodesic submanifolds

We estimate the $L^2$ norm of the restriction to a totally geodesic submanifold of the eigenfunctions of the Laplace-Beltrami operator on the standard flat torus $\mathbb{T}^d$, $d\ge2$. We reduce getting correct bounds to counting lattice points in the intersection of some $ν$-transverse bands on the sphere. Moreover, we prove the correct bounds for rational totally geodesic submanifolds of arbitrary codimension. In particular, we verify the conjecture of Bourgain-Rudnick on $L^2$-restriction estimates for rational hyperplanes. On $\mathbb{T}^2$, we prove the uniform $L^2$ restriction bounds for closed geodesics. On $\mathbb{T}^3$, we obtain explicit $L^2$ restriction estimates for the totally geodesic submanifolds, which improve the corresponding results by Burq-Gérard-Tzvetkov, Hu, Chen-Sogge.

math.AP↗

arXiv2019

Neuropathic Pain Diagnosis Simulator for Causal Discovery Algorithm Evaluation

Discovery of causal relations from observational data is essential for many disciplines of science and real-world applications. However, unlike other machine learning algorithms, whose development has been greatly fostered by a large amount of available benchmark datasets, causal discovery algorithms are notoriously difficult to be systematically evaluated because few datasets with known ground-truth causal relations are available. In this work, we handle the problem of evaluating causal discovery algorithms by building a flexible simulator in the medical setting. We develop a neuropathic pain diagnosis simulator, inspired by the fact that the biological processes of neuropathic pathophysiology are well studied with well-understood causal influences. Our simulator exploits the causal graph of the neuropathic pain pathology and its parameters in the generator are estimated from real-life patient cases. We show that the data generated from our simulator have similar statistics as real-world data. As a clear advantage, the simulator can produce infinite samples without jeopardizing the privacy of real-world patients. Our simulator provides a natural tool for evaluating various types of causal discovery algorithms, including those to deal with practical issues in causal discovery, such as unknown confounders, selection bias, and missing data. Using our simulator, we have evaluated extensively causal discovery algorithms under various settings.

cs.LG↗

arXiv2019

Generalization in Reinforcement Learning with Selective Noise Injection and Information Bottleneck

The ability for policies to generalize to new environments is key to the broad application of RL agents. A promising approach to prevent an agent's policy from overfitting to a limited set of training environments is to apply regularization techniques originally developed for supervised learning. However, there are stark differences between supervised learning and RL. We discuss those differences and propose modifications to existing regularization techniques in order to better adapt them to RL. In particular, we focus on regularization techniques relying on the injection of noise into the learned function, a family that includes some of the most widely used approaches such as Dropout and Batch Normalization. To adapt them to RL, we propose Selective Noise Injection (SNI), which maintains the regularizing effect the injected noise has, while mitigating the adverse effects it has on the gradient quality. Furthermore, we demonstrate that the Information Bottleneck (IB) is a particularly well suited regularization technique for RL as it is effective in the low-data regime encountered early on in training RL agents. Combining the IB with SNI, we significantly outperform current state of the art results, including on the recently proposed generalization benchmark Coinrun.

cs.LG↗

arXiv2019

Tightening Bounds for Variational Inference by Revisiting Perturbation Theory

Variational inference has become one of the most widely used methods in latent variable modeling. In its basic form, variational inference employs a fully factorized variational distribution and minimizes its KL divergence to the posterior. As the minimization can only be carried out approximately, this approximation induces a bias. In this paper, we revisit perturbation theory as a powerful way of improving the variational approximation. Perturbation theory relies on a form of Taylor expansion of the log marginal likelihood, vaguely in terms of the log ratio of the true posterior and its variational approximation. While first order terms give the classical variational bound, higher-order terms yield corrections that tighten it. However, traditional perturbation theory does not provide a lower bound, making it inapt for stochastic optimization. In this paper, we present a similar yet alternative way of deriving corrections to the ELBO that resemble perturbation theory, but that result in a valid bound. We show in experiments on Gaussian Processes and Variational Autoencoders that the new bounds are more mass covering, and that the resulting posterior covariances are closer to the true posterior and lead to higher likelihoods on held-out data.

stat.ML↗

arXiv2019

An Imprint of the Galactic Magnetic Field in the Diffuse Unpolarized Dust Emission

It is well known that aligned, aspherical dust grains emit polarized radiation and that the degree of polarization depends on the angle $ψ$ between the interstellar magnetic field and the line of sight. However, anisotropy of the dust absorption cross sections also modulates the $total\ intensity$ of the radiation as the viewing geometry changes. We report a detection of this effect in the high Galactic latitude $Planck$ data, finding that the 353 GHz dust intensity per $N_{\rm HI}$ is smaller when the Galactic magnetic field is mostly in the plane of the sky and larger when the field is mostly along the line of sight. These variations are of opposite sign and roughly equal magnitude as the changes in polarized intensity per $N_{\rm HI}$ with $ψ$, as predicted. In principle, the variation in intensity can be used in conjunction with the dust polarization angle to constrain the full 3D orientation of the Galactic magnetic field.

astro-ph.GA↗

arXiv2019

Interpretable Outcome Prediction with Sparse Bayesian Neural Networks in Intensive Care

Clinical decision making is challenging because of pathological complexity, as well as large amounts of heterogeneous data generated as part of routine clinical care. In recent years, machine learning tools have been developed to aid this process. Intensive care unit (ICU) admissions represent the most data dense and time-critical patient care episodes. In this context, prediction models may help clinicians determine which patients are most at risk and prioritize care. However, flexible tools such as artificial neural networks (ANNs) suffer from a lack of interpretability limiting their acceptability to clinicians. In this work, we propose a novel interpretable Bayesian neural network architecture which offers both the flexibility of ANNs and interpretability in terms of feature selection. In particular, we employ a sparsity inducing prior distribution in a tied manner to learn which features are important for outcome prediction. We evaluate our approach on the task of mortality prediction using two real-world ICU cohorts. In collaboration with clinicians we found that, in addition to the predicted outcome results, our approach can provide novel insights into the importance of different clinical measurements. This suggests that our model can support medical experts in their decision making process.

cs.LG↗

arXiv2019

Crystal electric field splitting and f-electron hybridization in heavy fermion CePt2In7

We use high-resolution angle-resolved photoemission spectroscopy to investigate the electronic structure of the antiferromagnetic heavy fermion compound CePt2In7, which is a member of the CeIn3-derived heavy fermion material family. Weak hybridization among 4f electron states and conduction bands was identified in CePt2In7 at low temperature much weaker than that in the other heavy fermion compounds like CeIrIn5 and CeRhIn5. The Ce 4f spectrum shows fine structures near the Fermi energy, reflecting the crystal electric field splitting of the 4f^1_5/2 and 4f^1_7/2 states. Also, we find that the Fermi surface has a strongly three-dimensional topology, in agreement with density-functional theory calculations.

cond-mat.str-el↗

arXiv2019

Generating an Overview Report over Many Documents

How to efficiently generate an accurate, well-structured overview report (ORPT) over thousands of related documents is challenging. A well-structured ORPT consists of sections of multiple levels (e.g., sections and subsections). None of the existing multi-document summarization (MDS) algorithms is directed toward this task. To overcome this obstacle, we present NDORGS (Numerous Documents' Overview Report Generation Scheme) that integrates text filtering, keyword scoring, single-document summarization (SDS), topic modeling, MDS, and title generation to generate a coherent, well-structured ORPT. We then devise a multi-criteria evaluation method using techniques of text mining and multi-attribute decision making on a combination of human judgments, running time, information coverage, and topic diversity. We evaluate ORPTs generated by NDORGS on two large corpora of documents, where one is classified and the other unclassified. We show that, using Saaty's pairwise comparison 9-point scale and under TOPSIS, the ORPTs generated on SDS's with the length of 20% of the original documents are the best overall on both datasets.

cs.CL↗