Searcharxiv⌕ Search

arXiv subjects

Peng Xu

Publications and source records attributed to Peng Xu.

At least 397 records · Page 22Linked to original sources
arXiv2018

Two-stage frequency recognition method based on correlated component analysis for SSVEP-based BCI

Canonical correlation analysis (CCA) is a state-of-the-art method for frequency recognition in steady-state visual evoked potential (SSVEP)-based brain-computer interface (BCI) systems. Various extended methods have been developed, and among such methods, a combination method of CCA and individual-template-based CCA (IT-CCA) has achieved excellent performance. However, CCA requires the canonical vectors to be orthogonal, which may not be a reasonable assumption for EEG analysis. In the current study, we propose using the correlated component analysis (CORRCA) rather than CCA to implement frequency recognition. CORRCA can relax the constraint of canonical vectors in CCA, and generate the same projection vector for two multichannel EEG signals. Furthermore, we propose a two-stage method based on the basic CORRCA method (termed TSCORRCA). Evaluated on a benchmark dataset of thirty-five subjects, the experimental results demonstrate that CORRCA significantly outperformed CCA, and TSCORRCA obtained the best performance among the compared methods. This study demonstrates that CORRCA-based methods have great potential for implementing high-performance SSVEP-based BCI systems.

q-bio.NC↗
arXiv2018

PlusEmo2Vec at SemEval-2018 Task 1: Exploiting emotion knowledge from emoji and #hashtags

This paper describes our system that has been submitted to SemEval-2018 Task 1: Affect in Tweets (AIT) to solve five subtasks. We focus on modeling both sentence and word level representations of emotion inside texts through large distantly labeled corpora with emojis and hashtags. We transfer the emotional knowledge by exploiting neural network models as feature extractors and use these representations for traditional machine learning models such as support vector regression (SVR) and logistic regression to solve the competition tasks. Our system is placed among the Top3 for all subtasks we participated.

cs.CL↗
arXiv2018

Coulomb blockade regions in sputter-deposited titanium nitride films

We present topographic and spectroscopic scanning tunneling microscopy measurements taken on a 21 nm thick TiN film at a temperature of 4.2 K -- above the superconducting transition temperature (T_c = 3.8 K) of the sample. The film was polycrystalline with crystallite diameters of d~19 nm, consistent with other films prepared under similar conditions. The spectroscopic maps show on average a shallow V-shape around V_b = 0 V consistent with a sample near the Mott insulation transition. In selected regions on several samples we additionally observed signs of Coulomb blockade. The corresponding peak structures are typically asymmetric with respect to bias voltage indicating coupling to two very different tunneling barriers. Furthermore, the peak structures appear with constant peak-peak spacing which indicates quantum dot states within the Coulomb blockade island. In this paper we discuss one such Coulomb blockade area and its implications in detail.

cond-mat.mes-hall↗
arXiv2018

Neural Fine-Grained Entity Type Classification with Hierarchy-Aware Loss

The task of Fine-grained Entity Type Classification (FETC) consists of assigning types from a hierarchy to entity mentions in text. Existing methods rely on distant supervision and are thus susceptible to noisy labels that can be out-of-context or overly-specific for the training sentence. Previous methods that attempt to address these issues do so with heuristics or with the help of hand-crafted features. Instead, we propose an end-to-end solution with a neural network model that uses a variant of cross- entropy loss function to handle out-of-context labels, and hierarchical loss normalization to cope with overly-specific ones. Also, previous work solve FETC a multi-label classification followed by ad-hoc post-processing. In contrast, our solution is more elegant: we use public word embeddings to train a single-label that jointly learns representations for entity mentions and their context. We show experimentally that our approach is robust against noise and consistently outperforms the state-of-the-art on established benchmarks for the task.

cs.CL↗
arXiv2018

SketchMate: Deep Hashing for Million-Scale Human Sketch Retrieval

We propose a deep hashing framework for sketch retrieval that, for the first time, works on a multi-million scale human sketch dataset. Leveraging on this large dataset, we explore a few sketch-specific traits that were otherwise under-studied in prior literature. Instead of following the conventional sketch recognition task, we introduce the novel problem of sketch hashing retrieval which is not only more challenging, but also offers a better testbed for large-scale sketch analysis, since: (i) more fine-grained sketch feature learning is required to accommodate the large variations in style and abstraction, and (ii) a compact binary code needs to be learned at the same time to enable efficient retrieval. Key to our network design is the embedding of unique characteristics of human sketch, where (i) a two-branch CNN-RNN architecture is adapted to explore the temporal ordering of strokes, and (ii) a novel hashing loss is specifically designed to accommodate both the temporal and abstract traits of sketches. By working with a 3.8M sketch dataset, we show that state-of-the-art hashing models specifically engineered for static images fail to perform well on temporal sketch data. Our network on the other hand not only offers the best retrieval performance on various code sizes, but also yields the best generalization performance under a zero-shot setting and when re-purposed for sketch recognition. Such superior performances effectively demonstrate the benefit of our sketch-specific design.

cs.CV↗
arXiv2018

Ferromagnetism and spin-dependent transport at a complex oxide interface

Complex oxide interfaces are a promising platform for studying a wide array of correlated electron phenomena in low-dimensions, including magnetism and superconductivity. The microscopic origin of these phenomena in complex oxide interfaces remains an open question. Here we investigate for the first time the magnetic properties of semi-insulating NdTiO$_3$/SrTiO$_3$ (NTO/STO) interfaces and present the first milli-Kelvin study of NTO/STO. The magnetoresistance (MR) reveals signatures of local ferromagnetic order and of spin-dependent thermally-activated transport, which are described quantitatively by a simple phenomenological model. We discuss possible origins of the interfacial ferromagnetism. In addition, the MR also shows transient hysteretic features on a timescale of ~10-100 seconds. We demonstrate that these are consistent with an extrinsic magneto-thermal origin, which may have been misinterpreted in previous reports of magnetism in STO-based oxide interfaces. The existence of these two MR regimes (steady-state and transient) highlights the importance of time-dependent measurements for distinguishing signatures of ferromagnetism from other effects that can produce hysteresis at low temperatures.

cond-mat.mes-hall↗
arXiv2018

On certain Iwahori representations of unramified $U(2, 1)$ in characteristic $p$

Let $F$ be a non-archimedean local field of odd residue characteristic $p$. Let $G$ be the unramified unitary group $U(2, 1)(E/F)$, and $K$ be a maximal compact open subgroup of $G$. For an $\overline{\mathbf{F}}_p$-smooth representation $π$ of $G$ containing a weight $σ$ of $K$, we follow the work of Hu (\cite{Hu12}) to attach $π$ a certain $I_K$-subrepresentation, where $I_K$ is the Iwahori subgroup in $K$. In terms of such an $I_K$-subrepresentation, we prove a sufficient condition for $π$ to be non-finitely presented. We determine such an $I_K$-subrepresentation explicitly, when $π$ is either a spherical universal Hecke module or an irreducible principal series.

math.RT↗
arXiv2018

Irreducible $p$-modular representations of unramified $U(2,1)$

Let $E/F$ be a unramified quadratic extension of non-archimedean local fields of odd characteristic $p$, and $G$ be the unramified unitary group $U(2, 1)(E/F)$. For an irreducible smooth representation $π$ of $G$ over $\overline{\mathbf{F}}_p$, with an underlying irreducible smooth representation $σ$ of a maximal compact open subgroup $K$, we prove that $π$ admits eigenvectors for an appropriate Hecke operator $T_σ$, and we classify those $π$ with non-zero eigenvalues for $T_σ$ by a tree argument; as a corollary, we show $π$ is supersingular if and only if it is supercuspidal.

math.RT↗
arXiv2018

Inexact Non-Convex Newton-Type Methods

For solving large-scale non-convex problems, we propose inexact variants of trust region and adaptive cubic regularization methods, which, to increase efficiency, incorporate various approximations. In particular, in addition to approximate sub-problem solves, both the Hessian and the gradient are suitably approximated. Using rather mild conditions on such approximations, we show that our proposed inexact methods achieve similar optimal worst-case iteration complexities as the exact counterparts. Our proposed algorithms, and their respective theoretical analysis, do not require knowledge of any unknowable problem-related quantities, and hence are easily implementable in practice. In the context of finite-sum problems, we then explore randomized sub-sampling methods as ways to construct the gradient and Hessian approximations and examine the empirical performance of our algorithms on some real datasets.

math.OC↗
arXiv2018

Second-Order Optimization for Non-Convex Machine Learning: An Empirical Study

While first-order optimization methods such as stochastic gradient descent (SGD) are popular in machine learning (ML), they come with well-known deficiencies, including relatively-slow convergence, sensitivity to the settings of hyper-parameters such as learning rate, stagnation at high training errors, and difficulty in escaping flat regions and saddle points. These issues are particularly acute in highly non-convex settings such as those arising in neural networks. Motivated by this, there has been recent interest in second-order methods that aim to alleviate these shortcomings by capturing curvature information. In this paper, we report detailed empirical evaluations of a class of Newton-type methods, namely sub-sampled variants of trust region (TR) and adaptive regularization with cubics (ARC) algorithms, for non-convex ML problems. In doing so, we demonstrate that these methods not only can be computationally competitive with hand-tuned SGD with momentum, obtaining comparable or better generalization performance, but also they are highly robust to hyper-parameter settings. Further, in contrast to SGD with momentum, we show that the manner in which these Newton-type methods employ curvature information allows them to seamlessly escape flat regions and saddle points.

math.OC↗
arXiv2018

Investigations on Knowledge Base Embedding for Relation Prediction and Extraction

We report an evaluation of the effectiveness of the existing knowledge base embedding models for relation prediction and for relation extraction on a wide range of benchmarks. We also describe a new benchmark, which is much larger and complex than previous ones, which we introduce to help validate the effectiveness of both tasks. The results demonstrate that knowledge base embedding models are generally effective for relation prediction but unable to give improvements for the state-of-art neural relation extraction model with the existing strategies, while pointing limitations of existing methods.

cs.CL↗
arXiv2018

Quantum repeater without Bell measurment in double quantum dot system

We propose a Bell measurement free scheme to implement a quantum repeater in GaAs/AlGa double qunatum dot systems.we prove the four pairs of double quantum dots compose an entanglement unit, given the the initial state is singlet states. Our shceme differs from the famous Duan-Lukin-Cirac-zoller(DLCZ) protocol in that Bell measurements are unneccessary for the entanglement swapping,which provides great advantages and conveniences in experimental implementaion. Our scheme significantly improve the success probability of quantum repeaters based on solid state quantum devices.

quant-ph↗
arXiv2017

High Fidelity Single-qubit Gates of a Single Neutral Atom in the Magic-Intensity Optical Dipole Trap

We demonstrate high fidelity single-qubit gate operation in a trapped single neutral atom. The atom is trapped in the recently invented magic-intensity optical dipole trap (MI-ODT) with more stable magnetic field. The MI-ODT efficiently mitigates the detrimental effects of light shifts thus sufficiently improves the performance of single qubit-gates. The gates are driven with microwave, and the fidelity of gate operation is characterized by using the randomized benchmarking method. We obtain an average error per Clifford gate of $3.0(7)\times10^{-5}$ which is much below the error threshold ($10^{-4}$) for fault-tolerance. This error is found to be dominated by qubit dephasing, and the corresponding coherence time relevant to the Clifford gates is also measured experimentally. This work is an essential step toward the construction of a scalable quantum computer with neutral atoms trapped in an MI-ODT array.

quant-ph↗
arXiv2017

Two-photon driven Kerr resonator for quantum annealing with three-dimensional circuit QED

We propose a realizable circuit QED architecture for engineering states of a superconducting resonator off-resonantly coupled to an ancillary superconducting qubit. The qubit-resonator dispersive interaction together with a microwave drive applied to the qubit gives rise to a Kerr resonator with two-photon driving that enables us to efficiently engineer the quantum state of the resonator such as generation of the Schrodinger cat states for resonator-based universal quantum computation. Moreover, the presented architecture is easily scalable for solving optimization problem mapped into the Ising spin glass model, and thus served as a platform for quantum annealing. Although various scalable architecture with superconducting qubits have been proposed for realizing quantum annealer, the existing annealers are currently limited to the coherent time of the qubits. Here, based on the protocol for realizing two-photon driven Kerr resonator in three-dimensional circuit QED (3D cQED), we propose a flexible and scalable hardware for implementing quantum annealer that combines the advantage of the long coherence times attainable in 3D cQED and the recently proposed resonator based Lechner-Hauke-Zoller (LHZ) scheme. In the proposed resonator based LHZ annealer, each spin is encoded in the subspace formed by two coherent state of 3D microwave superconducting resonator with opposite phase, and thus the fully-connected Ising model is mapped onto the network of the resonator with local tunable three-resonator interaction. This hardware architecture provides a promising physical platform for realizing quantum annealer with improved coherence.

quant-ph↗
arXiv2017

Investigation of the formation process of soap bubbles from soap film

In this paper, we study the formation process of a soap bubble by blowing soap film. Both bubble diameter and formation position were investigated in experiments. We found that the ratio between bubble size and soap film column is constant, and that the formation position increases linearly within a critical length-range. We used the theory of Rayleigh-Plateau instability to explain these findings. The theoretical explanations are consistent with the experimental data.

physics.flu-dyn↗
arXiv2017

Rebuilding of destroyed spin squeezing in noisy environments

We investigate the process of spin squeezing in a ferromagnetic dipolar spin-1 Bose-Einstein condensate under the driven oneaxis twisting scheme, with emphasis on the detrimental effect of noisy environments (stray magnetic fields) which completely destroy the spin squeezing. By applying concatenated dynamical decoupling pulse sequences with a moderate bias magnetic field to suppress the effect of the noisy environments, we faithfully reconstruct the spin squeezing process under realistic experimental conditions. Our noise-resistant method is ready to be employed to generate the spin squeezed state in a dipolar spin-1 Bose-Einstein condensate and paves a feasible way to the Heisenberg-limit quantum metrology

cond-mat.quant-gas↗
arXiv2017

Socratic Learning: Augmenting Generative Models to Incorporate Latent Subsets in Training Data

A challenge in training discriminative models like neural networks is obtaining enough labeled training data. Recent approaches use generative models to combine weak supervision sources, like user-defined heuristics or knowledge bases, to label training data. Prior work has explored learning accuracies for these sources even without ground truth labels, but they assume that a single accuracy parameter is sufficient to model the behavior of these sources over the entire training set. In particular, they fail to model latent subsets in the training data in which the supervision sources perform differently than on average. We present Socratic learning, a paradigm that uses feedback from a corresponding discriminative model to automatically identify these subsets and augments the structure of the generative model accordingly. Experimentally, we show that without any ground truth labels, the augmented generative model reduces error by up to 56.06% for a relation extraction task compared to a state-of-the-art weak supervision technique that utilizes generative models.

cs.LG↗