arXiv subjects

Jing Li

Publications and source records attributed to Jing Li.

At least 577 records · Page 32Linked to original sources

arXiv2021

Solving the Data Sparsity Problem in Predicting the Success of the Startups with Machine Learning Methods

Predicting the success of startup companies is of great importance for both startup companies and investors. It is difficult due to the lack of available data and appropriate general methods. With data platforms like Crunchbase aggregating the information of startup companies, it is possible to predict with machine learning algorithms. Existing research suffers from the data sparsity problem as most early-stage startup companies do not have much data available to the public. We try to leverage the recent algorithms to solve this problem. We investigate several machine learning algorithms with a large dataset from Crunchbase. The results suggest that LightGBM and XGBoost perform best and achieve 53.03% and 52.96% F1 scores. We interpret the predictions from the perspective of feature contribution. We construct portfolios based on the models and achieve high success rates. These findings have substantial implications on how machine learning methods can help startup companies and investors.

cs.LG↗

arXiv2021

Taming Overconfident Prediction on Unlabeled Data from Hindsight

Minimizing prediction uncertainty on unlabeled data is a key factor to achieve good performance in semi-supervised learning (SSL). The prediction uncertainty is typically expressed as the \emph{entropy} computed by the transformed probabilities in output space. Most existing works distill low-entropy prediction by either accepting the determining class (with the largest probability) as the true label or suppressing subtle predictions (with the smaller probabilities). Unarguably, these distillation strategies are usually heuristic and less informative for model training. From this discernment, this paper proposes a dual mechanism, named ADaptive Sharpening (\ADS), which first applies a soft-threshold to adaptively mask out determinate and negligible predictions, and then seamlessly sharpens the informed predictions, distilling certain predictions with the informed ones only. More importantly, we theoretically analyze the traits of \ADS by comparing with various distillation strategies. Numerous experiments verify that \ADS significantly improves the state-of-the-art SSL methods by making it a plug-in. Our proposed \ADS forges a cornerstone for future distillation-based SSL research.

cs.LG↗

arXiv2021

Global Classical Solutions to the Compressible Navier-Stokes Equations with Slip Boundary Conditions in 3D Exterior Domains

We are concerned with the global existence of classical solutions to the barotropic compressible Navier-Stokes equations with slip boundary condition in a three-dimensional (3D) exterior domain. We demonstrate that the classical solutions exists globally in time provided that the initial total energy is suitably small. It is worth noting that the initial density is allowed to have large oscillations and contain vacuum states. For our purpose, some new techniques and methods are adopted to obtain necessary a priori estimates, especially the estimates on the boundary. Moveover, we also give the large-time behavior of the classical solutions what we have gotten.

math.AP↗

arXiv2021

Determination of weak squeezed vacuum state through photon statistics measurement

Weak squeezed vacuum light, especially resonant to the atomic transition, plays an important role in quantum storage and generation of various quantum sources. However, the general homodyne detection (HD) cannot determine weak squeezing due to the low signal to noise ratio and the limited resolution of the HD system. Here we provide an alternative method based on photon statistics measurement to determine the weak squeezing of the squeezed vacuum light generated from an optical parametric oscillator working far below the threshold. The approach is established the relationship between the squeezing parameter and the second-order degree of coherence. The theoretical analysis agrees well with the experiment results. The advantage of this method is that it provides a feasible and reliable experimental measure to determine the weak squeezing with high precision and the measurement is independent on the detection efficiency. This method can be used to measure other quantum features for various quantum states with extremely weak non-classicality.

quant-ph↗

arXiv2021

Spatial Curvature and Large Scale Lorentz Violation

The tension between the Hubble constant obtained from the local measurements and from cosmic microwave background (CMB) measurements motivated us to consider the cosmological model beyond $Λ$CDM one. We investigate the cosmology in the large scale Lorentz violation model with non-vanishing spatial curvature. The degeneracy among spatial curvature, cosmological constant and cosmological contortion distribution makes the model viable in describing the known observation date. We get some constraints on the spatial curvature by the comparison of the relation between measured distance modulus and red-shift with the predicted one, the evolution of matter density over time and the evolution of effective cosmological constant. The performance of large scale Lorentz violation model with non-vanishing spatial curvature under these constrains is discussed.

gr-qc↗

arXiv2021

Fragmenting Active Asteroid 331P/Gibbs

We describe active asteroid 331P/Gibbs (2012 F5) using archival Hubble Space Telescope data taken between 2015 and 2018. 331P is an outer main-belt active asteroid with a long-lived debris trail that formed in 2011. Embedded in the debris trail we identify 19 fragments with radii between 0.04 and 0.11 km (albedo 0.05 assumed) containing about 1 percent of the mass of the primary nucleus. The largest shows a photometric range (1.5 magnitudes), a V-shaped minimum and a two-peaked lightcurve period near 9 hours, consistent with a symmetric contact binary (Drahus and Waniak 2016). Less convincing explanations are that 331P-A is a monolithic, elongated splinter or that its surface shows hemispheric 4:1 albedo variations. The debris trail is composed of centimeter sized and larger particles ejected with characteristic 10 cm s$^{-1}$ speeds following a size distribution with index q = 3.7+/-0.1 to 4.1+/-0.2. The HST data show that earlier, ground-based measurements of the nucleus were contaminated by near-nucleus debris, which cleared by 2015. We find that the primary nucleus has effective radius 0.8+/-0.1 km and is in rapid rotation (3.26+/-0.01 hour, range 0.25 magnitudes, minimum density 1600 kg/m3 if strengthless. The properties of 331P are consistent with a) formation about 1.5 Myr ago by impact shattering of a precursor body (Novakovic et al. 2014) b) spin-up by radiation torques to critical rotation c) ejection of about 1 percent of the nucleus mass in mid-2011 by rotational instability and d) subsequent evolution of the fragments and dispersal of the debris by radiation pressure.

astro-ph.EP↗

arXiv2021

Gate-tunable artificial nucleus in graphene

We report an atomically-precise integration of individual nitrogen (N) dopant as an in-plane artificial nucleus in a graphene device by atomic implantation to probe its gate-tunable quantum states and correlation effects. The N dopant creates the characteristic resonance state in the conduction band, revealing a giant carrier-dependent energetic renormalization up to 350 meV with respect to the Dirac point, accompanied by the observation of long-range screening effects. Joint density functional theory and tight-binding calculations with modified perturbation potential corroborate experimental findings and highlight the short-range character of N-induced perturbation.

cond-mat.mtrl-sci↗

arXiv2021

Electronic and Thermal Properties of $\text{GeTe/Sb}_{2}\text{Te}_{3}$ Superlattices by ab-initio Approach: Impact of Van der Waals Gaps on Vertical Lattice Thermal Conductivity

In the last decade, several works have focused on exploring the material and electrical properties of $\text{GeTe/Sb}_{2}\text{Te}_{3}$ superlattices (SLs) in particular because of some first device implementations demonstrating interesting performances such as fast switching speed, low energy consumption, and non-volatility. However, the switching mechanism in such SL-based devices remains under debate. In this work, we investigate the prototype $\text{GeTe/Sb}_{2}\text{Te}_{3}$ SLs, to analyze fundamentally their electronic and thermal properties by ab initio methods. We find that the resistive contrast is small among the different phases of $\text{GeTe/Sb}_{2}\text{Te}_{3}$ because of a small electronic gap (about 0.1 eV) and a consequent semi-metallic-like behavior. At the same time the out-of-plane lattice thermal conductivity is rather small, while varying up to four times among the different phases, from 0.11 to 0.45 W/m$^{-1}$K$^{-1}$, intimately related to the number of Van der Waals (VdW) gaps in a unit block. Such findings confirm the importance of the thermal improvement achievable in $\text{GeTe/Sb}_{2}\text{Te}_{3}$ super-lattices devices, highlighting the impact of the material stacking and the role of VdW gaps on the thermal engineering of the Phase-Change Memory cell.

cond-mat.mtrl-sci↗

arXiv2021

Non-reciprocal Pauli Spin Blockade in a Silicon Double Quantum Dot

Spin qubits in gate-defined silicon quantum dots are receiving increased attention thanks to their potential for large-scale quantum computing. Readout of such spin qubits is done most accurately and scalably via Pauli spin blockade (PSB), however various mechanisms may lift PSB and complicate readout. In this work, we present an experimental observation of a new, highly prevalent PSB-lifting mechanism in a silicon double quantum dot due to incoherent tunneling between different spin manifolds. Through dispersively-detected magnetospectroscopy of the double quantum dot in 16 charge configurations, we find the mechanism to be energy-level selective and non-reciprocal for neighbouring charge configurations. Additionally, using input-output theory we report a large coupling of different electron spin manifolds of 7.90 $μ$eV, the largest reported to date, indicating an enhanced spin-orbit coupling which may enable all-electrical qubit control.

cond-mat.mes-hall↗

arXiv2021

Proxy-bridged Image Reconstruction Network for Anomaly Detection in Medical Images

Anomaly detection in medical images refers to the identification of abnormal images with only normal images in the training set. Most existing methods solve this problem with a self-reconstruction framework, which tends to learn an identity mapping and reduces the sensitivity to anomalies. To mitigate this problem, in this paper, we propose a novel Proxy-bridged Image Reconstruction Network (ProxyAno) for anomaly detection in medical images. Specifically, we use an intermediate proxy to bridge the input image and the reconstructed image. We study different proxy types, and we find that the superpixel-image (SI) is the best one. We set all pixels' intensities within each superpixel as their average intensity, and denote this image as SI. The proposed ProxyAno consists of two modules, a Proxy Extraction Module and an Image Reconstruction Module. In the Proxy Extraction Module, a memory is introduced to memorize the feature correspondence for normal image to its corresponding SI, while the memorized correspondence does not apply to the abnormal images, which leads to the information loss for abnormal image and facilitates the anomaly detection. In the Image Reconstruction Module, we map an SI to its reconstructed image. Further, we crop a patch from the image and paste it on the normal SI to mimic the anomalies, and enforce the network to reconstruct the normal image even with the pseudo abnormal SI. In this way, our network enlarges the reconstruction error for anomalies. Extensive experiments on brain MR images, retinal OCT images and retinal fundus images verify the effectiveness of our method for both image-level and pixel-level anomaly detection.

eess.IV↗

arXiv2021

Sub-angstrom Non-invasive Imaging of Atomic Arrangement in 2D Hybrid Perovskites

Non-invasive imaging of the atomic arrangement in two-dimensional (2D) Ruddlesden-Popper hybrid Perovskites (RPPs), as well as understanding the related effects is challenging, due to the insulating nature and softness of the organic layers which also obscure the underlying inorganic lattice. Here, we demonstrate a sub-angstrom resolution imaging of both soft organic layers and inorganic framework in a prototypical 2D lead-halide RPP crystal via combined tip-functionalized Scanning Tunneling Microscopy (STM) and non-contact Atomic Force Microscope (ncAFM) corroborated by theoretical simulations. STM measurements unveil the atomic reconstruction of the inorganic lead-halide lattice and overall twin-domain composition of the RPP crystal, while ncAFM measurements with a CO-tip enable non-perturbative visualization of the cooperative reordering of surface organic cations driven by their hydrogen bonding interactions with the inorganic lattice. Moreover, such a joint technique also allows for the atomic-scale imaging of the electrostatic potential variation across the twin-domain walls, revealing alternating quasi-one-dimensional (1D) electron and hole-channels at neighboring twin-boundaries, which may influence in-plane exciton transport and dissociation.

cond-mat.mtrl-sci↗

arXiv2021

Close Latency--Security Trade-off for the Nakamoto Consensus

Bitcoin is a peer-to-peer electronic cash system invented by Nakamoto in 2008. While it has attracted much research interest, its exact latency and security properties remain open. Existing analyses provide security and latency (or confirmation time) guarantees that are too loose for practical use. In fact the best known upper bounds are several orders of magnitude larger than a lower bound due to a well-known private-mining attack. This paper describes a continuous-time model for blockchains and develops a rigorous analysis that yields close upper and lower bounds for the latency--security trade-off. For example, when the adversary controls 10\% of the total mining power and the block propagation delays are within 10 seconds, a Bitcoin block is secured with less than $10^{-3}$ error probability if it is confirmed after four hours, or with less than $10^{-9}$ error probability if confirmed after ten hours. These confirmation times are about two hours away from their corresponding lower bounds. To establish such close bounds, the blockchain security question is reduced to a race between the Poisson adversarial mining process and a renewal process formed by a certain species of honest blocks. The moment generation functions of relevant renewal times are derived in closed form. The general formulas from the analysis are then applied to study the latency--security trade-off of several well-known proof-of-work longest-chain cryptocurrencies. Guidance is also provided on how to set parameters for different purposes.

cs.CR↗

arXiv2021

Phyllotaxis-inspired Nanosieves with Multiplexed Orbital Angular Momentum

Nanophotonic platforms such as metasurfaces, achieving arbitrary phase profiles within ultrathin thickness, emerge as miniaturized, ultracompact and kaleidoscopic optical vortex generators. However, it is often required to segment or interleave independent subarray metasurfaces to multiplex optical vortices in a single nano device, which in turn affects the compactness and channel capacity of the device. Here, inspired by phyllotaxis patterns in pine cones and sunflowers, we theoretically prove and experimentally report that multiple optical vortices can be produced in a single compact phyllotaxis nanosieve, both in free space and on a chip, where one metaatom may contribute to many vortices simultaneously. The time resolved dynamics of on chip interference wavefronts between multiple plasmonic vortices was revealed by ultrafast time-resolved photoemission electron microscopy. Our nature inspired optical vortex generator would facilitate various vortex related optical applications, including structured wavefront shaping, free space and plasmonic vortices, and high capacity information metaphotonics.

physics.optics↗

arXiv2021

Revisiting 3D ResNets for Video Recognition

A recent work from Bello shows that training and scaling strategies may be more significant than model architectures for visual recognition. This short note studies effective training and scaling strategies for video recognition models. We propose a simple scaling strategy for 3D ResNets, in combination with improved training strategies and minor architectural changes. The resulting models, termed 3D ResNet-RS, attain competitive performance of 81.0 on Kinetics-400 and 83.8 on Kinetics-600 without pre-training. When pre-trained on a large Web Video Text dataset, our best model achieves 83.5 and 84.3 on Kinetics-400 and Kinetics-600. The proposed scaling rule is further evaluated in a self-supervised setup using contrastive learning, demonstrating improved performance. Code is available at: https://github.com/tensorflow/models/tree/master/official.

cs.CV↗

arXiv2021

Minority Oversampling for Imbalanced Time Series Classification

Many important real-world applications involve time-series data with skewed distribution. Compared to conventional imbalance learning problems, the classification of imbalanced time-series data is more challenging due to high dimensionality and high inter-variable correlation. This paper proposes a structure preserving Oversampling method to combat the High-dimensional Imbalanced Time-series classification (OHIT). OHIT first leverages a density-ratio based shared nearest neighbor clustering algorithm to capture the modes of minority class in high-dimensional space. It then for each mode applies the shrinkage technique of large-dimensional covariance matrix to obtain accurate and reliable covariance structure. Finally, OHIT generates the structure-preserving synthetic samples based on multivariate Gaussian distribution by using the estimated covariance matrices. Experimental results on several publicly available time-series datasets (including unimodal and multimodal) demonstrate the superiority of OHIT against the state-of-the-art oversampling algorithms in terms of F1, G-mean, and AUC.

cs.LG↗

arXiv2021

Fuzzy linear systems and core-EP inverses

The main purpose of this paper is to provide a solution of the consistent fuzzy linear system and a generalized solution of the inconsistent fuzzy linear system involving the core-EP inverse of an associated matrix. Before this can be achieved, it is necessary to study the block structure of the core-EP inverse. Finally, results are illustrated with some numerical examples.

math.RA↗

arXiv2021

Audio2Gestures: Generating Diverse Gestures from Speech Audio with Conditional Variational Autoencoders

Generating conversational gestures from speech audio is challenging due to the inherent one-to-many mapping between audio and body motions. Conventional CNNs/RNNs assume one-to-one mapping, and thus tend to predict the average of all possible target motions, resulting in plain/boring motions during inference. In order to overcome this problem, we propose a novel conditional variational autoencoder (VAE) that explicitly models one-to-many audio-to-motion mapping by splitting the cross-modal latent code into shared code and motion-specific code. The shared code mainly models the strong correlation between audio and motion (such as the synchronized audio and motion beats), while the motion-specific code captures diverse motion information independent of the audio. However, splitting the latent code into two parts poses training difficulties for the VAE model. A mapping network facilitating random sampling along with other techniques including relaxed motion loss, bicycle constraint, and diversity loss are designed to better train the VAE. Experiments on both 3D and 2D motion datasets verify that our method generates more realistic and diverse motions than state-of-the-art methods, quantitatively and qualitatively. Finally, we demonstrate that our method can be readily used to generate motion sequences with user-specified motion clips on the timeline. Code and more results are at https://jingli513.github.io/audio2gestures.

cs.CV↗

arXiv2021

OLR 2021 Challenge: Datasets, Rules and Baselines

This paper introduces the sixth Oriental Language Recognition (OLR) 2021 Challenge, which intends to improve the performance of language recognition systems and speech recognition systems within multilingual scenarios. The data profile, four tasks, two baselines, and the evaluation principles are introduced in this paper. In addition to the Language Identification (LID) tasks, multilingual Automatic Speech Recognition (ASR) tasks are introduced to OLR 2021 Challenge for the first time. The challenge this year focuses on more practical and challenging problems, with four tasks: (1) constrained LID, (2) unconstrained LID, (3) constrained multilingual ASR, (4) unconstrained multilingual ASR. Baselines for LID tasks and multilingual ASR tasks are provided, respectively. The LID baseline system is an extended TDNN x-vector model constructed with Pytorch. A transformer-based end-to-end model is provided as the multilingual ASR baseline system. These recipes will be online published, and available for participants to construct their own LID or ASR systems. The baseline results demonstrate that those tasks are rather challenging and deserve more effort to achieve better performance.

cs.CL↗