arXiv subjects

Liu Yang

Publications and source records attributed to Liu Yang.

At least 199 records · Page 11Linked to original sources

arXiv2020

APTER: Aggregated Prognosis Through Exponential Reweighting

This paper considers the task of learning how to make a prognosis of a patient based on his/her micro-array expression levels. The method is an application of the aggregation method as recently proposed in the literature on theoretical machine learning, and excels in its computational convenience and capability to deal with high-dimensional data. A formal analysis of the method is given, yielding rates of convergence similar to what traditional techniques obtain, while it is shown to cope well with an exponentially large set of features. Those results are supported by numerical simulations on a range of publicly available survival-micro-array datasets. It is empirically found that the proposed technique combined with a recently proposed preprocessing technique gives excellent performances.

stat.ME↗

arXiv2020

IART: Intent-aware Response Ranking with Transformers in Information-seeking Conversation Systems

Personal assistant systems, such as Apple Siri, Google Assistant, Amazon Alexa, and Microsoft Cortana, are becoming ever more widely used. Understanding user intent such as clarification questions, potential answers and user feedback in information-seeking conversations is critical for retrieving good responses. In this paper, we analyze user intent patterns in information-seeking conversations and propose an intent-aware neural response ranking model "IART", which refers to "Intent-Aware Ranking with Transformers". IART is built on top of the integration of user intent modeling and language representation learning with the Transformer architecture, which relies entirely on a self-attention mechanism instead of recurrent nets. It incorporates intent-aware utterance attention to derive an importance weighting scheme of utterances in conversation context with the aim of better conversation history understanding. We conduct extensive experiments with three information-seeking conversation data sets including both standard benchmarks and commercial data. Our proposed model outperforms all baseline methods with respect to a variety of metrics. We also perform case studies and analysis of learned user intent and its impact on response ranking in information-seeking conversations to provide interpretation of results.

cs.IR↗

arXiv2019

Reflection Phase Shift of One-dimensional Plasmon Polaritons in Carbon Nanotubes

We investigated, both experimentally and theoretically, the reflection phase shift (RPS) of one-dimensional plasmon polaritons. We launched 1D plasmon polaritons in carbon nanotube and probed the plasmon interference pattern using scanning near-field optical microscopy (SNOM) technique, through which a non-zero phase shift was observed. We further developed a theory to understand the nonzero phase shift of 1D polaritons, and found that the RPS can be understood by considering the evanescent field beyond the nanotube end. Interesting, our theory shows a strong dependence of RPS on polaritons wavelength and nanotube diameter, which is in stark contrast to 2D plasmon polaritons in graphene where the RPS is a constant. In short wave region, the RPS of 1D polaritons only depends on a dimensionless variable -- the ratio between polaritons wavelength and nanotube diameter. These results provide fundamental insights into the reflection of polaritons in 1D system, and could facilitate the design of ultrasmall 1D polaritonic devices, such as resonators, interferometers.

cond-mat.mes-hall↗

arXiv2019

Surrogate Losses in Passive and Active Learning

Active learning is a type of sequential design for supervised machine learning, in which the learning algorithm sequentially requests the labels of selected instances from a large pool of unlabeled data points. The objective is to produce a classifier of relatively low risk, as measured under the 0-1 loss, ideally using fewer label requests than the number of random labeled data points sufficient to achieve the same. This work investigates the potential uses of surrogate loss functions in the context of active learning. Specifically, it presents an active learning algorithm based on an arbitrary classification-calibrated surrogate loss function, along with an analysis of the number of label requests sufficient for the classifier returned by the algorithm to achieve a given risk under the 0-1 loss. Interestingly, these results cannot be obtained by simply optimizing the surrogate risk via active learning to an extent sufficient to provide a guarantee on the 0-1 loss, as is common practice in the analysis of surrogate losses for passive learning. Some of the results have additional implications for the use of surrogate losses in passive learning.

math.ST↗

arXiv2019

A GMM based algorithm to generate point-cloud and its application to neuroimaging

Recent years have witnessed the emergence of 3D medical imaging techniques with the development of 3D sensors and technology. Due to the presence of noise in image acquisition, registration researchers focused on an alternative way to represent medical images. An alternative way to analyze medical imaging is by understanding the 3D shapes represented in terms of point-cloud. Though in the medical imaging community, 3D point-cloud processing is not a ``go-to'' choice, it is a ``natural'' way to capture 3D shapes. However, as the number of samples for medical images are small, researchers have used pre-trained models to fine-tune on medical images. Furthermore, due to different modality in medical images, standard generative models can not be used to generate new samples of medical images. In this work, we use the advantage of point-cloud representation of 3D structures of medical images and propose a Gaussian mixture model-based generation scheme. Our proposed method is robust to outliers. Experimental validation has been performed to show that the proposed scheme can generate new 3D structures using interpolation techniques, i.e., given two 3D structures represented as point-clouds, we can generate point-clouds in between. We have also generated new point-clouds for subjects with and without dementia and show that the generated samples are indeed closely matched to the respective training samples from the same class.

cs.LG↗

arXiv2019

An "augmentation-free" rotation invariant classification scheme on point-cloud and its application to neuroimaging

Recent years have witnessed the emergence and increasing popularity of 3D medical imaging techniques with the development of 3D sensors and technology. However, achieving geometric invariance in the processing of 3D medical images is computationally expensive but nonetheless essential due to the presence of possible errors caused by rigid registration techniques. An alternative way to analyze medical imaging is by understanding the 3D shapes represented in terms of point-cloud. Though in the medical imaging community, 3D point-cloud processing is not a "go-to" choice, it is a canonical way to preserve rotation invariance. Unfortunately, due to the presence of discrete topology, one can not use the standard convolution operator on point-cloud. To the best of our knowledge, the existing ways to do "convolution" can not preserve the rotation invariance without explicit data augmentation. Therefore, we propose a rotation invariant convolution operator by inducing topology from hypersphere. Experimental validation has been performed on publicly available OASIS dataset in terms of classification accuracy between subjects with (without) dementia, demonstrating the usefulness of our proposed method in terms of model complexity, classification accuracy, and last but most important invariance to rotations.

eess.IV↗

arXiv2019

Computing and Testing Small Connectivity in Near-Linear Time and Queries via Fast Local Cut Algorithms

Consider the following "local" cut-detection problem in a directed graph: We are given a seed vertex $x$ and need to remove at most $k$ edges so that at most $ν$ edges can be reached from $x$ (a "local" cut) or output $\bot$ to indicate that no such cut exists. If we are given query access to the input graph, then this problem can in principle be solved without reading the whole graph and with query complexity depending on $k$ and $ν$. In this paper we consider a slack variant of this problem where, when such a cut exists, we can output a cut with up to $O(kν)$ edges reachable from $x$. We present a simple randomized algorithm spending $O(k^2ν)$ time and $O(kν)$ queries for the above variant, improving in particular a previous time bound of $O(k^{O(k)}ν)$ by Chechik et al. [SODA '17]. We also extend our algorithm to handle an approximate variant. We demonstrate that these local algorithms are versatile primitives for designing substantially improved algorithms for classic graph problems by providing the following three applications. (Throughout, $\tilde O(T)$ hides $\operatorname{polylog}(T)$.) (1) A randomized algorithm for the classic $k$-vertex connectivity problem that takes near-linear time when $k=O(\operatorname{polylog}(n))$, namely $\tilde O(m+nk^3)$ time in undirected graphs. For directed graphs our $\tilde O(mk^2)$-time algorithm is near-linear when $k=O(\operatorname{polylog}(n))$. Our techniques also yield an improved approximation scheme. (2) Property testing algorithms for $k$-edge and -vertex connectivity with query complexities that are near-linear in $k$, exponentially improving the state-of-the-art. This resolves two open problems, one by Goldreich and Ron [STOC '97] and one by Orenstein and Ron [Theor. Comput Sci. '11]. (3) A faster algorithm for computing the maximal $k$-edge connected subgraphs, improving prior work of Chechik et al. [SODA '17].

cs.DS↗

arXiv2019

POIRot: A rotation invariant omni-directional pointnet

Point-cloud is an efficient way to represent 3D world. Analysis of point-cloud deals with understanding the underlying 3D geometric structure. But due to the lack of smooth topology, and hence the lack of neighborhood structure, standard correlation can not be directly applied on point-cloud. One of the popular approaches to do point correlation is to partition the point-cloud into voxels and extract features using standard 3D correlation. But this approach suffers from sparsity of point-cloud and hence results in multiple empty voxels. One possible solution to deal with this problem is to learn a MLP to map a point or its local neighborhood to a high dimensional feature space. All these methods suffer from a large number of parameters requirement and are susceptible to random rotations. A popular way to make the model "invariant" to rotations is to use data augmentation techniques with small rotations but the potential drawback includes \item more training samples \item susceptible to large rotations. In this work, we develop a rotation invariant point-cloud segmentation and classification scheme based on the omni-directional camera model (dubbed as {\bf POIRot$^1$}). Our proposed model is rotationally invariant and can preserve geometric shape of a 3D point-cloud. Because of the inherent rotation invariant property, our proposed framework requires fewer number of parameters (please see \cite{Iandola2017SqueezeNetAA} and the references therein for motivation of lean models). Several experiments have been performed to show that our proposed method can beat the state-of-the-art algorithms in classification and part segmentation applications.

cs.CV↗

arXiv2019

Highly-scalable, physics-informed GANs for learning solutions of stochastic PDEs

Uncertainty quantification for forward and inverse problems is a central challenge across physical and biomedical disciplines. We address this challenge for the problem of modeling subsurface flow at the Hanford Site by combining stochastic computational models with observational data using physics-informed GAN models. The geographic extent, spatial heterogeneity, and multiple correlation length scales of the Hanford Site require training a computationally intensive GAN model to thousands of dimensions. We develop a hierarchical scheme for exploiting domain parallelism, map discriminators and generators to multiple GPUs, and employ efficient communication schemes to ensure training stability and convergence. We developed a highly optimized implementation of this scheme that scales to 27,500 NVIDIA Volta GPUs and 4584 nodes on the Summit supercomputer with a 93.1% scaling efficiency, achieving peak and sustained half-precision rates of 1228 PF/s and 1207 PF/s.

physics.comp-ph↗

arXiv2019

BERT with History Answer Embedding for Conversational Question Answering

Conversational search is an emerging topic in the information retrieval community. One of the major challenges to multi-turn conversational search is to model the conversation history to answer the current question. Existing methods either prepend history turns to the current question or use complicated attention mechanisms to model the history. We propose a conceptually simple yet highly effective approach referred to as history answer embedding. It enables seamless integration of conversation history into a conversational question answering (ConvQA) model built on BERT (Bidirectional Encoder Representations from Transformers). We first explain our view that ConvQA is a simplified but concrete setting of conversational search, and then we provide a general framework to solve ConvQA. We further demonstrate the effectiveness of our approach under this framework. Finally, we analyze the impact of different numbers of history turns under different settings to provide new insights into conversation history modeling in ConvQA.

cs.IR↗

arXiv2019

Perfect Optical Nonreciprocity in a Double-Cavity Optomechanical System

Nonreciprocal devices are indispensable for building quantum networks and ubiquitous in modern communication technology. Here, we use optomechanical interaction and linearly-coupled interaction to realize optical nonreciprocal transmission in a double-cavity optomechanical system. The scheme relies on the interference between the two interactions. We derive the essential conditions to realize perfect optical nonreciprocity in the system, and analyse the properties of optical nonreciprocal transmission and the output fields from mechanical mode. These results can be used to control optical transmission in quantum information processing.

quant-ph↗

arXiv2019

Potential Flow Generator with $L_2$ Optimal Transport Regularity for Generative Models

We propose a potential flow generator with $L_2$ optimal transport regularity, which can be easily integrated into a wide range of generative models including different versions of GANs and flow-based models. We show the correctness and robustness of the potential flow generator in several 2D problems, and illustrate the concept of "proximity" due to the $L_2$ optimal transport regularity. Subsequently, we demonstrate the effectiveness of the potential flow generator in image translation tasks with unpaired training data from the MNIST dataset and the CelebA dataset.

cs.LG↗

arXiv2019

Attentive History Selection for Conversational Question Answering

Conversational question answering (ConvQA) is a simplified but concrete setting of conversational search. One of its major challenges is to leverage the conversation history to understand and answer the current question. In this work, we propose a novel solution for ConvQA that involves three aspects. First, we propose a positional history answer embedding method to encode conversation history with position information using BERT in a natural way. BERT is a powerful technique for text representation. Second, we design a history attention mechanism (HAM) to conduct a "soft selection" for conversation histories. This method attends to history turns with different weights based on how helpful they are on answering the current question. Third, in addition to handling conversation history, we take advantage of multi-task learning (MTL) to do answer prediction along with another essential conversation task (dialog act prediction) using a uniform model architecture. MTL is able to learn more expressive and generic representations to improve the performance of ConvQA. We demonstrate the effectiveness of our model with extensive experimental evaluations on QuAC, a large-scale ConvQA dataset. We show that position information plays an important role in conversation history modeling. We also visualize the history attention and provide new insights into conversation history understanding.

cs.IR↗

arXiv2019

A Hybrid Retrieval-Generation Neural Conversation Model

Intelligent personal assistant systems that are able to have multi-turn conversations with human users are becoming increasingly popular. Most previous research has been focused on using either retrieval-based or generation-based methods to develop such systems. Retrieval-based methods have the advantage of returning fluent and informative responses with great diversity. However, the performance of the methods is limited by the size of the response repository. On the other hand, generation-based methods can produce highly coherent responses on any topics. But the generated responses are often generic and not informative due to the lack of grounding knowledge. In this paper, we propose a hybrid neural conversation model that combines the merits of both response retrieval and generation methods. Experimental results on Twitter and Foursquare data show that the proposed model outperforms both retrieval-based methods and generation-based methods (including a recently proposed knowledge-grounded neural conversation model) under both automatic evaluation metrics and human evaluation. We hope that the findings in this study provide new insights on how to integrate text retrieval and text generation models for building conversation systems.

cs.IR↗

arXiv2019

Stochastic Polynomial Optimization

This paper studies stochastic optimization problems with polynomials. We propose an optimization model with sample averages and perturbations. The Lasserre type Moment-SOS relaxations are used to solve the sample average optimization. Properties of the optimization and its relaxations are studied. Numerical experiments are presented.

math.OC↗

arXiv2019

Transmission Nonreciprocity in a Mutually Coupled Circulating Structure

Breaking Lorentz reciprocity was believed to be a prerequisite for nonreciprocal transmissions of light fields, so the possibility of nonreciprocity by linear optical systems was mostly ignored. We put forward a structure of three mutually coupled microcavities or optical fiber rings to realize optical nonreciprocity. Although its couplings with the fields from two different input ports are constantly equal, such system transmits them nonreciprocally either under the saturation of an optical gain in one of the cavities or with the asymmetric couplings of the circulating fields in different cavities. The structure made up of optical fiber rings can perform nonreciprocal transmissions as a time-independent linear system without breaking Lorentz reciprocity. Optical isolation for inputs simultaneously from two different ports and even approximate optical isolator operations are implementable with the structure.

physics.optics↗

arXiv2019

Identify the stiffness of DNA via deep learning

DNA detection is of great significance in the point-of-care diagnostics. The stiffness of DNA, varying with its sequence and mechanochemical environment, could be a potential marker for DNA identification. The steric configurations of DNA fragments with different stiffness were simulated by employing the Kirchhoff theory of thin elastic rods. We identified the stiffness of DNA with the trained convolutional neural network on the simulated image set. The identification accuracy reached 99.85%. The stiffness-based identification provided a promising approach for DNA detection.

physics.bio-ph↗

arXiv2019

A Deep Look into Neural Ranking Models for Information Retrieval

Ranking models lie at the heart of research on information retrieval (IR). During the past decades, different techniques have been proposed for constructing ranking models, from traditional heuristic methods, probabilistic methods, to modern machine learning methods. Recently, with the advance of deep learning technology, we have witnessed a growing body of work in applying shallow or deep neural networks to the ranking problem in IR, referred to as neural ranking models in this paper. The power of neural ranking models lies in the ability to learn from the raw text inputs for the ranking problem to avoid many limitations of hand-crafted features. Neural networks have sufficient capacity to model complicated tasks, which is needed to handle the complexity of relevance estimation in ranking. Since there have been a large variety of neural ranking models proposed, we believe it is the right time to summarize the current status, learn from existing methodologies, and gain some insights for future development. In contrast to existing reviews, in this survey, we will take a deep look into the neural ranking models from different dimensions to analyze their underlying assumptions, major design principles, and learning strategies. We compare these models through benchmark tasks to obtain a comprehensive empirical understanding of the existing techniques. We will also discuss what is missing in the current literature and what are the promising and desired future directions.

cs.IR↗