arXiv subjects

Yuhao Zhang

Publications and source records attributed to Yuhao Zhang.

180 records · Page 10Linked to original sources

arXiv2019

Multi-Perspective Relevance Matching with Hierarchical ConvNets for Social Media Search

Despite substantial interest in applications of neural networks to information retrieval, neural ranking models have only been applied to standard ad hoc retrieval tasks over web pages and newswire documents. This paper proposes MP-HCNN (Multi-Perspective Hierarchical Convolutional Neural Network) a novel neural ranking model specifically designed for ranking short social media posts. We identify document length, informal language, and heterogeneous relevance signals as features that distinguish documents in our domain, and present a model specifically designed with these characteristics in mind. Our model uses hierarchical convolutional layers to learn latent semantic soft-match relevance signals at the character, word, and phrase levels. A pooling-based similarity measurement layer integrates evidence from multiple types of matches between the query, the social media post, as well as URLs contained in the post. Extensive experiments using Twitter data from the TREC Microblog Tracks 2011--2014 show that our model significantly outperforms prior feature-based as well and existing neural ranking models. To our best knowledge, this paper presents the first substantial work tackling search over social media posts using neural ranking models.

cs.IR↗

arXiv2019

Universal Dependency Parsing from Scratch

This paper describes Stanford's system at the CoNLL 2018 UD Shared Task. We introduce a complete neural pipeline system that takes raw text as input, and performs all tasks required by the shared task, ranging from tokenization and sentence segmentation, to POS tagging and dependency parsing. Our single system submission achieved very competitive performance on big treebanks. Moreover, after fixing an unfortunate bug, our corrected system would have placed the 2nd, 1st, and 3rd on the official evaluation metrics LAS,MLAS, and BLEX, and would have outperformed all submission systems on low-resource treebank categories on all metrics by a large margin. We further show the effectiveness of different model components through extensive ablation studies.

cs.CL↗

arXiv2018

Tight Competitive Ratios of Classic Matching Algorithms in the Fully Online Model

Huang et al.~(STOC 2018) introduced the fully online matching problem, a generalization of the classic online bipartite matching problem in that it allows all vertices to arrive online and considers general graphs. They showed that the ranking algorithm by Karp et al.~(STOC 1990) is strictly better than $0.5$-competitive and the problem is strictly harder than the online bipartite matching problem in that no algorithms can be $(1-1/e)$-competitive. This paper pins down two tight competitive ratios of classic algorithms for the fully online matching problem. For the fractional version of the problem, we show that a natural instantiation of the water-filling algorithm is $2-\sqrt{2} \approx 0.585$-competitive, together with a matching hardness result. Interestingly, our hardness result applies to arbitrary algorithms in the edge-arrival models of the online matching problem, improving the state-of-art $\frac{1}{1+\ln 2} \approx 0.5906$ upper bound. For integral algorithms, we show a tight competitive ratio of $\approx 0.567$ for the ranking algorithm on bipartite graphs, matching a hardness result by Huang et al. (STOC 2018).

cs.DS↗

arXiv2018

Cross-type Biomedical Named Entity Recognition with Deep Multi-Task Learning

Motivation: State-of-the-art biomedical named entity recognition (BioNER) systems often require handcrafted features specific to each entity type, such as genes, chemicals and diseases. Although recent studies explored using neural network models for BioNER to free experts from manual feature engineering, the performance remains limited by the available training data for each entity type. Results: We propose a multi-task learning framework for BioNER to collectively use the training data of different types of entities and improve the performance on each of them. In experiments on 15 benchmark BioNER datasets, our multi-task model achieves substantially better performance compared with state-of-the-art BioNER systems and baseline neural sequence labeling models. Further analysis shows that the large performance gains come from sharing character- and word-level information among relevant biomedical entities across differently labeled corpora.

cs.IR↗

arXiv2018

Learning to Summarize Radiology Findings

The Impression section of a radiology report summarizes crucial radiology findings in natural language and plays a central role in communicating these findings to physicians. However, the process of generating impressions by summarizing findings is time-consuming for radiologists and prone to errors. We propose to automate the generation of radiology impressions with neural sequence-to-sequence learning. We further propose a customized neural model for this task which learns to encode the study background information and use this information to guide the decoding process. On a large dataset of radiology reports collected from actual hospital studies, our model outperforms existing non-neural and neural baselines under the ROUGE metrics. In a blind experiment, a board-certified radiologist indicated that 67% of sampled system summaries are at least as good as the corresponding human-written summaries, suggesting significant clinical validity. To our knowledge our work represents the first attempt in this direction.

cs.CL↗

arXiv2018

Graph Convolution over Pruned Dependency Trees Improves Relation Extraction

Dependency trees help relation extraction models capture long-range relations between words. However, existing dependency-based models either neglect crucial information (e.g., negation) by pruning the dependency trees too aggressively, or are computationally inefficient because it is difficult to parallelize over different tree structures. We propose an extension of graph convolutional networks that is tailored for relation extraction, which pools information over arbitrary dependency structures efficiently in parallel. To incorporate relevant information while maximally removing irrelevant content, we further apply a novel pruning strategy to the input trees by keeping words immediately around the shortest path between the two entities among which a relation might hold. The resulting model achieves state-of-the-art performance on the large-scale TACRED dataset, outperforming existing sequence and dependency-based neural models. We also show through detailed analysis that this model has complementary strengths to sequence models, and combining them further improves the state of the art.

cs.CL↗

arXiv2018

Online Makespan Minimization: The Power of Restart

We consider the online makespan minimization problem on identical machines. Chen and Vestjens (ORL 1997) show that the largest processing time first (LPT) algorithm is 1.5-competitive. For the special case of two machines, Noga and Seiden (TCS 2001) introduce the SLEEPY algorithm that achieves a competitive ratio of $(5 - \sqrt{5})/2 \approx 1.382$, matching the lower bound by Chen and Vestjens (ORL 1997). Furthermore, Noga and Seiden note that in many applications one can kill a job and restart it later, and they leave an open problem whether algorithms with restart can obtain better competitive ratios. We resolve this long-standing open problem on the positive end. Our algorithm has a natural rule for killing a processing job: a newly-arrived job replaces the smallest processing job if 1) the new job is larger than other pending jobs, 2) the new job is much larger than the processing one, and 3) the processed portion is small relative to the size of the new job. With appropriate choice of parameters, we show that our algorithm improves the 1.5 competitive ratio for the general case, and the 1.382 competitive ratio for the two-machine case.

cs.DS↗

arXiv2018

Energy-Efficient Transmission of Hybrid Array with Non-Ideal Power Amplifiers and Circuitry

This paper presents a new approach to efficiently maximizing the energy efficiency (EE) of hybrid arrays under a practical setting of non-ideal power amplifiers (PAs) and non-negligible circuit power, where coherent and non-coherent beamforming are considered. As a key contribution, we reveal that a bursty transmission mode can be energy-efficient to achieve steady transmissions of a data stream under the practical setting. This is distinctively different from existing studies under ideal circuits and PAs, where continuous transmissions are the most energy-efficient. Another important contribution is that the optimal transmit duration and powers are identified to balance energy consumptions in the non-ideal circuits and PAs, and maximize the EE. This is achieved by establishing the most energy-efficient structure of transmit powers, given a transmit duration, and correspondingly partitioning the non-convex feasible region of the transmit duration into segments with self-contained convexity or concavity. Evident from simulations, significant EE gains of the proposed approach are demonstrated through comparisons with the state of the art, and the superiority of the bursty transmission mode is confirmed especially under low data rate demands.

cs.IT↗

arXiv2018

How to Match when All Vertices Arrive Online

We introduce a fully online model of maximum cardinality matching in which all vertices arrive online. On the arrival of a vertex, its incident edges to previously-arrived vertices are revealed. Each vertex has a deadline that is after all its neighbors' arrivals. If a vertex remains unmatched until its deadline, the algorithm must then irrevocably either match it to an unmatched neighbor, or leave it unmatched. The model generalizes the existing one-sided online model and is motivated by applications including ride-sharing platforms, real-estate agency, etc. We show that the Ranking algorithm by Karp et al. (STOC 1990) is $0.5211$-competitive in our fully online model for general graphs. Our analysis brings a novel charging mechanic into the randomized primal dual technique by Devanur et al. (SODA 2013), allowing a vertex other than the two endpoints of a matched edge to share the gain. To our knowledge, this is the first analysis of Ranking that beats $0.5$ on general graphs in an online matching problem, a first step towards solving the open problem by Karp et al. (STOC 1990) about the optimality of Ranking on general graphs. If the graph is bipartite, we show that the competitive ratio of Ranking is between $0.5541$ and $0.5671$. Finally, we prove that the fully online model is strictly harder than the previous model as no online algorithm can be $0.6317 < 1-\frac{1}{e}$-competitive in our model even for bipartite graphs.

cs.DS↗

arXiv2018

Energy Efficiency Maximization for CoMP Joint Transmission with Non-Ideal Power Amplifiers

Coordinated multipoint (CoMP) joint transmission (JT) can save a great deal of energy especially for cell-edge users due to strengthened received signal, but at the cost of deploying and coordinating cooperative nodes, which degrades energy efficiency (EE), particularly when considerable amount of energy is consumed by nonideal hardware circuit. In this paper, we study energyefficient cooperation establishment, including cooperative nodes selection (CNS) and power allocation, to satisfy a required data rate in coherent JT-CoMP networks with non-ideal power amplifiers (PAs) and circuit power. The selection priority lemma is proved first, and then the formulated discrete combinatorial EE optimization is resolved by proposing node selection criterion and deriving closedform expressions of optimal transmission power. Therefore, an efficient algorithm is provided and its superiority is validated by Monte Carlo simulations, which also show the effects of non-ideal PA and the data rate demand on EE and optimal number of active nodes.

cs.IT↗

arXiv2018

Energy-efficient resource allocation for hybrid bursty services in multi-relay OFDM networks

In this paper, we propose Mapped Two-way Water Filling (MTWF) scheme to maximize Energy Efficiency (EE) for hybrid bursty services with Quality of Services (QoS) requirements in Two-Way Multi-Relay (TWMR) OFDM networks. The bursty traffic is first analyzed by strictly proved equivalent homogeneous Poisson process, based on which the QoS requirements are converted into sum-rate constraints. The formulated non-convex EE maximization problem, including subcarrier assignment, Relay Selection (RS) and rate allocation, is NP-hard involving combinatorial optimization. To conduct optimal RS on each subcarrier without priori bursty traffic knowledge, we utilize some approximate relationships under high data rate demands to remove its dependence on two-way data rates, and simplify the whole optimization problem as well. After the optimal channel configuration is obtained, which only depends on channel conditions, subcarrier assignment is attained through Elitist Selection Genetic Algorithm (ESGA), and rate allocation of each service is fulfilled by deducing two-way water filling principle. A new equivalent optimization objective function is proposed next as the simple evaluating index in ESGA to reduce complexity. Finally, simulations are carried out to verify the superiority and convergence of our scheme, as well as the applicability for different scenarios.

cs.IT↗

arXiv2018

Energy Efficiency Analysis of Heterogeneous Cellular Networks With Extra Cell Range Expansion

The split control and user plane is key to the future heterogeneous cellular network (HCN), where the small cells are dedicated for the most data transmission while the macrocells are mainly responsible for the control signaling. Adapting to this technology, we propose a general and tractable framework of extra cell range expansion (CRE) by introducing an additional bias factor to enlarge the range of small cells flexibly for the extra offloaded macrousers in a two-tier HCN, where the macrocell and small cell users have different required data rates. Using stochastic geometry, we analyze the energy efficiency (EE) of the extra CRE with joint low power transmission and resource partitioning, where the coverages of EE and data rate are formulated theoretically. Numerical simulations verify that the proposed extra CRE can improve the EE performance of HCN, and also show that deploying more small cells can provide benefits for EE coverage, but the EE improvement becomes saturated if the small cell density exceeds a threshold. Instead of establishing the detail configuration, this paper can provide some valuable insights and guidelines to the practical design of future networks, especially for the traffic offloading in HCN.

cs.IT↗

arXiv2018

Energy Efficiency Maximization of Full-Duplex Two-Way Relay With Non-Ideal Power Amplifiers and Non-Negligible Circuit Power

In this paper, we maximize the energy efficiency (EE) of full-duplex (FD) two-way relay (TWR) systems under non-ideal power amplifiers (PAs) and non-negligible transmission-dependent circuit power. We start with the case where only the relay operates full duplex and two timeslots are required for TWR. Then, we extend to the advanced case, where the relay and the two nodes all operate full duplex, and accomplish TWR in a single timeslot. In both cases, we establish the intrinsic connections between the optimal transmit powers and durations, based on which the original non-convex EE maximization can be convexified and optimally solved. Simulations show the superiority of FD-TWR in terms of EE, especially when traffic demand is high. The simulations also reveal that the maximum EE of FD-TWR is more sensitive to the PA efficiency, than it is to self-cancellation. The full FD design of FD-TWR is susceptible to traffic imbalance, while the design with only the relay operating in the FD mode exhibits strong tolerance.

cs.IT↗

arXiv2018

Optimal Pilot Symbols Ratio in terms of Spectrum and Energy Efficiency in Uplink CoMP Networks

In wireless networks, Spectrum Efficiency (SE) and Energy Efficiency (EE) can be affected by the channel estimation that needs to be well designed in practice. In this paper, considering channel estimation error and non-ideal backhaul links, we optimize the pilot symbols ratio in terms of SE and EE in uplink Coordinated Multi-point (CoMP) networks. Modeling the channel estimation error, we formulate the SE and EE maximization problems by analyzing the system capacity with imperfect channel estimation. The maximal system capacity in SE optimization and the minimal transmit power in EE optimization, which both have the closed-form expressions, are derived by some reasonable approximations to reduce the complexity of solving complicated equations. Simulations are carried out to validate the superiority of our scheme, verify the accuracy of our approximation, and show the effect of pilot symbols ratio.

cs.IT↗

arXiv2018

Energy-Efficient User Access Control and Resource Allocation in HCNs with Non-Ideal Circuitry

In this paper, we study the energy-efficient user access control (UAC) based on resource allocation (RA) in heterogeneous cellular networks (HCNs) with the required downlink data rate under non-ideal power amplifiers (PAs) and circuit power. It is proved that the energy consumption minimization is achieved when the typical user accesses only one base station (BS), while the other BSs remain in idle mode on the transmission resource allocated to this user. For this purpose, we reformulate the original non-convex optimization problem into a series of convex optimization problems where, in each case, the transmit power and duration of the accessed BS are determined. Then, the BS with the minimal energy consumption is selected for transmission. Considering the approximate situation, it is showed that the optimal transmit duration of the accessed BS can be estimated in closed form. The benefits of our proposed UAC and RA schemes are validated using numerical simulations, which also characterize the effect that non-ideal PAs have on the total energy consumption of different transmission schemes.

cs.IT↗

arXiv2017

Lie-Type Derivations of Nest Algebras on Banach Spaces

Let $\mathcal{X}$ be a Banach space over the complex field $\mathbb{C}$ and $\mathcal{B(X)}$ be the algebra of all bounded linear operators on $\mathcal{X}$. Let $\mathcal{N}$ be a non-trivial nest on $\mathcal{X}$, ${\rm Alg}\mathcal{N}$ be the nest algebra associated with $\mathcal{N}$, and $L\colon {\rm Alg}\mathcal{N}\longrightarrow \mathcal{B(X)}$ be a linear mapping. Suppose that $p_n(x_1,x_2,\cdots,x_n)$ is an $(n-1)$-th commutator defined by $n$ indeterminates $x_1, x_2, \cdots, x_n$. It is shown that $L$ satisfies the rule $$ L(p_n(A_1, A_2, \cdots, A_n))=\sum_{k=1}^{n}p_n(A_1, \cdots, A_{k-1}, L(A_k), A_{k+1}, \cdots, A_n) $$ for all $A_1, A_2, \cdots, A_n\in {\rm Alg}\mathcal{N}$ if and only if there exist a linear derivation $D\colon {\rm Alg}\mathcal{N}\longrightarrow \mathcal{B(X)}$ and a linear mapping $H\colon {\rm Alg}\mathcal{N}\longrightarrow \mathbb{C}I$ vanishing on each $(n-1)$-th commutator $p_n(A_1,A_2,\cdots, A_n)$ for all $A_1, A_2, \cdots, A_n\in {\rm Alg}\mathcal{N}$ such that $L(A)=D(A)+H(A)$ for all $A\in {\rm Alg}\mathcal{N}$.

math.FA↗

arXiv2016

Segmental Convolutional Neural Networks for Detection of Cardiac Abnormality With Noisy Heart Sound Recordings

Heart diseases constitute a global health burden, and the problem is exacerbated by the error-prone nature of listening to and interpreting heart sounds. This motivates the development of automated classification to screen for abnormal heart sounds. Existing machine learning-based systems achieve accurate classification of heart sound recordings but rely on expert features that have not been thoroughly evaluated on noisy recordings. Here we propose a segmental convolutional neural network architecture that achieves automatic feature learning from noisy heart sound recordings. Our experiments show that our best model, trained on noisy recording segments acquired with an existing hidden semi-markov model-based approach, attains a classification accuracy of 87.5% on the 2016 PhysioNet/CinC Challenge dataset, compared to the 84.6% accuracy of the state-of-the-art statistical classifier trained and evaluated on the same dataset. Our results indicate the potential of using neural network-based methods to increase the accuracy of automated classification of heart sound recordings for improved screening of heart diseases.

cs.SD↗

arXiv2011

Ultrathin MgB2 films fabricated on Al2O3 substrate by hybrid physical-chemical vapor deposition with high Tc and Jc

Ultrathin MgB2 superconducting films with a thickness down to 7.5 nm are epitaxially grown on (0001) Al2O3 substrate by hybrid physical-chemical vapor deposition method. The films are phase-pure, oxidation-free and continuous. The 7.5 nm thin film shows a Tc(0) of 34 K, which is so far the highest Tc(0) reported in MgB2 with the same thickness. The critical current density of ultrathin MgB2 films below 10 nm is demonstrated for the first time as Jc ~ 10^6 A cm^{-2} for the above 7.5 nm sample at 16 K. Our results reveal the excellent superconducting properties of ultrathin MgB2 films with thicknesses between 7.5 and 40 nm on Al2O3 substrate.

cond-mat.supr-con↗