BZPEERReading, trust and collaboration for research
Menu

Discover

GraphFollow connections around a paper, topic or saved view.

Workspaces

HomePick up where your research flow left off.InboxHandle requests, replies and notifications from one place.CollectionsCurate reading lists, evidence sets and shareable dossiers.ResearchSee your private provenance graph, trail and imports.CollaborationMove from discovery into focused discussion rooms.PublishDraft, preview and publish full-text research articles.

Network

ExpertsFind specialists worth following or asking for help.CommunitiesJoin research circles organized around shared work.PartnersSee external organizations active around the same topics.

Opportunities

ListingsBrowse roles, calls and collaborations with clearer fit signals.

Account

Log inReturn to your reading and collaboration workspace.Create accountStart saving work, following people and joining teams.
Log inCreate account

Researcher profile

Yingjie Li

Yingjie Li contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
astro-ph.GAastro-ph.EPastro-ph.SREmerging Technologiesgr-qcMachine Learningmath.STMethodologyphysics.opticsStatistics Theory

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author
6works
0followers
10topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

6 published item(s)

preprint2022arXiv

Local Spiral Structure Traced by Red Clump Stars

Using the cross-matched data of Gaia EDR3 and the 2MASS Point Source Catalog, a sample of RC stars with parallax accuracies better than 20% is identified and used to reveal the nearby spiral pattern traced by old stars. As shown in the overdensity distribution of RC stars, there is an arc-like feature extended from $l~\sim$ 90$^\circ$ to $\sim$ 243$^\circ$, which passed close to the Sun. This feature is probably an arm segment traced by old stars, indicating the Galaxy potential in the vicinity of the Sun. By comparing to the spiral arms depicted by young objects, we found that there are considerable offsets between the two different components of Galactic spiral arms. The spiral arm traced by RC stars tends to have a larger pitch angle, hence a more loose wound pattern.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Physics-aware Complex-valued Adversarial Machine Learning in Reconfigurable Diffractive All-optical Neural Network

Diffractive optical neural networks have shown promising advantages over electronic circuits for accelerating modern machine learning (ML) algorithms. However, it is challenging to achieve fully programmable all-optical implementation and rapid hardware deployment. Furthermore, understanding the threat of adversarial ML in such system becomes crucial for real-world applications, which remains unexplored. Here, we demonstrate a large-scale, cost-effective, complex-valued, and reconfigurable diffractive all-optical neural networks system in the visible range based on cascaded transmissive twisted nematic liquid crystal spatial light modulators. With the assist of categorical reparameterization, we create a physics-aware training framework for the fast and accurate deployment of computer-trained models onto optical hardware. Furthermore, we theoretically analyze and experimentally demonstrate physics-aware adversarial attacks onto the system, which are generated from a complex-valued gradient-based algorithm. The detailed adversarial robustness comparison with conventional multiple layer perceptrons and convolutional neural networks features a distinct statistical adversarial property in diffractive optical neural networks. Our full stack of software and hardware provides new opportunities of employing diffractive optics in a variety of ML tasks and enabling the research on optical adversarial ML.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

QuickSkill: Novice Skill Estimation in Online Multiplayer Games

Matchmaking systems are vital for creating fair matches in online multiplayer games, which directly affects players' satisfactions and game experience. Most of the matchmaking systems largely rely on precise estimation of players' game skills to construct equitable games. However, the skill rating of a novice is usually inaccurate, as current matchmaking rating algorithms require considerable amount of games for learning the true skill of a new player. Using these unreliable skill scores at early stages for matchmaking usually leads to disparities in terms of team performance, which causes negative game experience. This is known as the ''cold-start'' problem for matchmaking rating algorithms. To overcome this conundrum, this paper proposes QuickSKill, a deep learning based novice skill estimation framework to quickly probe abilities of new players in online multiplayer games. QuickSKill extracts sequential performance features from initial few games of a player to predict his/her future skill rating with a dedicated neural network, thus delivering accurate skill estimation at the player's early game stage. By employing QuickSKill for matchmaking, game fairness can be dramatically improved in the initial cold-start period. We conduct experiments in a popular mobile multiplayer game in both offline and online scenarios. Results obtained with two real-world anonymized gaming datasets demonstrate that proposed QuickSKill delivers precise estimation of game skills for novices, leading to significantly lower team skill disparities and better player game experience. To the best of our knowledge, proposed QuickSKill is the first framework that tackles the cold-start problem for traditional skill rating algorithms.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

Light Deflection under the Gravitational Field of Jupiter -- Testing General Relativity

We measured the relative positions between two pairs of compact extragalactic sources (CESs), J1925-2219 \& J1923-2104 (C1--C2) and J1925-2219 \& J1928-2035 (C1--C3) on 2020 October 23--25 and 2021 February 5 (totaling four epochs), respectively, using the Very Long Baseline Array (VLBA) at 15 GHz. Accounting for the deflection angle dominated by Jupiter, as well as the contributions from the Sun, planets other than Earth, the Moon and Ganymede (the most massive of the solar system's moons), our theoretical calculations predict that the dynamical ranges of the relative positions across four epochs in R.A. of the C1--C2 pair and C1--C3 pair are 841.2 and 1127.9 $μ$as, respectively. The formal accuracy in R.A. is about 20 $μ$as, but the error in Decl. is poor. The measured standard deviations of the relative positions across the four epochs are 51.0 and 29.7 $μ$as in R.A. for C1--C2 and C1--C3, respectively. These values indicate that the accuracy of the post-Newtonian relativistic parameter, $γ$, is $\sim 0.061$ for C1--C2 and $\sim 0.026$ for C1--C3. Combining the two CES pairs, the measured value of $γ$ is $0.984 \pm 0.037$, which is comparable to the latest published results for Jupiter as a gravitational lens reported by Fomalont \& Kopeikin, i.e., $1.01 \pm 0.03$.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

High Dimensional Classification for Spatially Dependent Data with Application to Neuroimaging

Discriminating patients with Alzheimer's disease (AD) from healthy subjects is a crucial task in the research of Alzheimer's disease. The task can be potentially achieved by linear discriminant analysis (LDA), which is one of the most classical and popular classification techniques. However, the classification problem becomes challenging for LDA because of the high-dimensionally and the spatial dependency of the brain imaging data. To address the challenges, researchers have proposed various ways to generalize LDA into high-dimensional context in recent years. However, these existing methods did not reach any consensus on how to incorporate spatially dependent structure. In light of the current needs and limitations, we propose a new classification method, named as Penalized Maximum Likelihood Estimation LDA (PMLE-LDA). The proposed method uses $Mat\acute{e}rn$ covariance function to describe the spatial correlation of brain regions. Additionally, PMLE is designed to model the sparsity of high-dimensional features. The spatial location information is used to address the singularity of the covariance. Tapering technique is introduced to reduce computational burden. We show in theory that the proposed method can not only provide consistent results of parameter estimation and feature selection, but also generate an asymptotically optimal classifier driven by high dimensional data with specific spatially dependent structure. Finally, the method is validated through simulations and an application into ADNI data for classifying Alzheimer's patients.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Searching for Molecular Outflows with Support Vector Machines: Dark Cloud Complex in Cygnus

We present a survey of molecular outflows across the dark cloud complex in the Cygnus region, based on 46.75 deg^2 field of CO isotopologues data from Milky Way Imaging Scroll Painting (MWISP) survey. A supervised machine learning algorithm, Support Vector Machine (SVM), is introduced to accelerate our visual assessment of outflow features in the data cube of 12CO and 13CO J = 1-0 emission. A total of 130 outflow candidates are identified, of which 77 show bipolar structures and 118 are new detections. Spatially, these outflows are located inside dense molecular clouds and some of them are found in clusters or in elongated linear structures tracing the underlying gas filament morphology. Along the line of sight, 97, 31, and 2 candidates reside in the Local, Perseus, and Outer arm, respectively. Young stellar objects as outflow drivers are found near most outflows, while 36 candidates show no associated source. The clusters of outflows that we detect are inhomogeneous in their properties; nevertheless, we show that the outflows cannot inject turbulent energy on cloud scales. Instead, at best, they are restricted to affecting the so called "clump" and "core" scales, and this only on short (~0.3 Myr) estimated timescales. Combined with outflow samples in the literature, our work shows a tight outflow mass-size correlation.

0 citations0 reviewsNo code linkedOpen access