arXiv subjects

Yifei Zhao

Publications and source records attributed to Yifei Zhao.

52 records · Page 3Linked to original sources

arXiv2022

$p$-torsion étale sheaves on the Jacobian of a curve

Suppose $X$ is a smooth, proper, geometrically connected curve over $\mathbb F_q$ with an $\mathbb F_q$-rational point $x_0$. For any $\mathbb F_q^{\times}$-character $σ$ of $π_1(X)$ trivial on $x_0$, we construct a functor $\mathbb L_n^σ$ from the derived category of coherent sheaves on the moduli space of deformations of $σ$ over the Witt ring $W_n(\mathbb F_q)$ to the derived category of constructible $W_n(\mathbb F_q)$-sheaves on the Jacobian of $X$. The functors $\mathbb L_n^σ$ categorify the Artin reciprocity map for geometric class field theory with $p$-torsion coefficients. We then give a criterion for the fully faithfulness of (an enhanced version of) $\mathbb L_n^σ$ in terms of the Hasse-Witt matrix of $X$.

math.AG↗

arXiv2022

RumorLens: Interactive Analysis and Validation of Suspected Rumors on Social Media

With the development of social media, various rumors can be easily spread on the Internet and such rumors can have serious negative effects on society. Thus, it has become a critical task for social media platforms to deal with suspected rumors. However, due to the lack of effective tools, it is often difficult for platform administrators to analyze and validate rumors from a large volume of information on a social media platform efficiently. We have worked closely with social media platform administrators for four months to summarize their requirements of identifying and analyzing rumors, and further proposed an interactive visual analytics system, RumorLens, to help them deal with the rumor efficiently and gain an in-depth understanding of the patterns of rumor spreading. RumorLens integrates natural language processing (NLP) and other data processing techniques with visualization techniques to facilitate interactive analysis and validation of suspected rumors. We propose well-coordinated visualizations to provide users with three levels of details of suspected rumors: an overview displays both spatial distribution and temporal evolution of suspected rumors; a projection view leverages a metaphor-based glyph to represent each suspected rumor and further enable users to gain a quick understanding of their overall characteristics and similarity with each other; a propagation view visualizes the dynamic spreading details of a suspected rumor with a novel circular visualization design, and facilitates interactive analysis and validation of rumors in a compact manner. By using a real-world dataset collected from Sina Weibo, one case study with a domain expert is conducted to evaluate

cs.HC↗

arXiv2022

Exploit Customer Life-time Value with Memoryless Experiments

As a measure of the long-term contribution produced by customers in a service or product relationship, life-time value, or LTV, can more comprehensively find the optimal strategy for service delivery. However, it is challenging to accurately abstract the LTV scene, model it reasonably, and find the optimal solution. The current theories either cannot precisely express LTV because of the single modeling structure, or there is no efficient solution. We propose a general LTV modeling method, which solves the problem that customers' long-term contribution is difficult to quantify while existing methods, such as modeling the click-through rate, only pursue the short-term contribution. At the same time, we also propose a fast dynamic programming solution based on a mutated bisection method and the memoryless repeated experiments assumption. The model and method can be applied to different service scenarios, such as the recommendation system. Experiments on real-world datasets confirm the effectiveness of the proposed model and optimization method. In addition, this whole LTV structure was deployed at a large E-commerce mobile phone application, where it managed to select optimal push message sending time and achieved a 10\% LTV improvement.

cs.AI↗

arXiv2021

Attributed Graph Neural Networks for Recommendation Systems on Large-Scale and Sparse Graph

Link prediction in structured-data is an important problem for many applications, especially for recommendation systems. Existing methods focus on how to learn the node representation based on graph-based structure. High-dimensional sparse edge features are not fully exploited. Because balancing precision and computation efficiency is significant for recommendation systems in real world, multiple-level feature representation in large-scale sparse graph still lacks effective and efficient solution. In this paper, we propose a practical solution about graph neural networks called Attributed Graph Convolutional Networks(AGCN) to incorporate edge attributes when apply graph neural networks in large-scale sparse networks. We formulate the link prediction problem as a subgraph classification problem. We firstly propose an efficient two-level projection to decompose topological structures to node-edge pairs and project them into the same interaction feature space. Then we apply multi-layer GCN to combine the projected node-edge pairs to capture the topological structures. Finally, the pooling representation of two units is treated as the input of classifier to predict the probability. We conduct offline experiments on two industrial datasets and one public dataset and demonstrate that AGCN outperforms other excellent baselines. Moreover, we also deploy AGCN method to important scenarios on Xianyu and AliExpress. In online systems, AGCN achieves over 5% improvement on online metrics.

cs.SI↗

arXiv2020

Smart obervation method with wide field small aperture telescopes for real time transient detection

Wide field small aperture telescopes (WFSATs) are commonly used for fast sky survey. Telescope arrays composed by several WFSATs are capable to scan sky several times per night. Huge amount of data would be obtained by them and these data need to be processed immediately. In this paper, we propose ARGUS (Astronomical taRGets detection framework for Unified telescopes) for real-time transit detection. The ARGUS uses a deep learning based astronomical detection algorithm implemented in embedded devices in each WFSATs to detect astronomical targets. The position and probability of a detection being an astronomical targets will be sent to a trained ensemble learning algorithm to output information of celestial sources. After matching these sources with star catalog, ARGUS will directly output type and positions of transient candidates. We use simulated data to test the performance of ARGUS and find that ARGUS can increase the performance of WFSATs in transient detection tasks robustly.

astro-ph.IM↗

arXiv2020

Semi-Supervised Learning for In-Game Expert-Level Music-to-Dance Translation

Music-to-dance translation is a brand-new and powerful feature in recent role-playing games. Players can now let their characters dance along with specified music clips and even generate fan-made dance videos. Previous works of this topic consider music-to-dance as a supervised motion generation problem based on time-series data. However, these methods suffer from limited training data pairs and the degradation of movements. This paper provides a new perspective for this task where we re-formulate the translation problem as a piece-wise dance phrase retrieval problem based on the choreography theory. With such a design, players are allowed to further edit the dance movements on top of our generation while other regression based methods ignore such user interactivity. Considering that the dance motion capture is an expensive and time-consuming procedure which requires the assistance of professional dancers, we train our method under a semi-supervised learning framework with a large unlabeled dataset (20x than labeled data) collected. A co-ascent mechanism is introduced to improve the robustness of our network. Using this unlabeled dataset, we also introduce self-supervised pre-training so that the translator can understand the melody, rhythm, and other components of music phrases. We show that the pre-training significantly improves the translation accuracy than that of training from scratch. Experimental results suggest that our method not only generalizes well over various styles of music but also succeeds in expert-level choreography for game players.

cs.CV↗

arXiv2020

Maximizing Cumulative User Engagement in Sequential Recommendation: An Online Optimization Perspective

To maximize cumulative user engagement (e.g. cumulative clicks) in sequential recommendation, it is often needed to tradeoff two potentially conflicting objectives, that is, pursuing higher immediate user engagement (e.g., click-through rate) and encouraging user browsing (i.e., more items exposured). Existing works often study these two tasks separately, thus tend to result in sub-optimal results. In this paper, we study this problem from an online optimization perspective, and propose a flexible and practical framework to explicitly tradeoff longer user browsing length and high immediate user engagement. Specifically, by considering items as actions, user's requests as states and user leaving as an absorbing state, we formulate each user's behavior as a personalized Markov decision process (MDP), and the problem of maximizing cumulative user engagement is reduced to a stochastic shortest path (SSP) problem. Meanwhile, with immediate user engagement and quit probability estimation, it is shown that the SSP problem can be efficiently solved via dynamic programming. Experiments on real-world datasets demonstrate the effectiveness of the proposed approach. Moreover, this approach is deployed at a large E-commerce platform, achieved over 7% improvement of cumulative clicks.

cs.IR↗

arXiv2020

Extensions by $\mathbf K_2$ and factorization line bundles

Let $X$ be a smooth, geometrically connected curve over a perfect field $k$. Given a connected, reductive group $G$, we prove that central extensions of $G$ by the sheaf $\mathbf K_2$ on the big Zariski site of $X$, studied by J.-L. Brylinski and P. Deligne, are equivalent to factorization line bundles on the Beilinson-Drinfeld affine Grassmannian $\operatorname{Gr}_G$. Our result affirms a conjecture of D. Gaitsgory and S. Lysenko and classifies factorization line bundles on $\operatorname{Gr}_G$.

math.AG↗

arXiv2020

Tame twistings and $Θ$-data

The goal of this paper is to assign an intrinsic meaning to the space of quantum parameters $\operatorname{Par}_G$ appearing in the geometric Langlands program of Beilinson-Drinfeld. We introduce tame twistings, a variant of twisted differential operators (TDOs) for which regularity of twisted $\mathscr D$-modules is well-defined. Our main result is that for a proper curve $X$, $\operatorname{Par}_G$ is precisely the moduli space of factorization tame twistings on the affine Grassmannian.

math.AG↗

arXiv2020

A Class D Power Amplifier for Multi-Frequency Eddy Current Testing Based on Multi-Simultaneous-Frequency Selective Harmonic Elimination Pulse Width Modulation

Efficiency and multisimultaneous-frequency (MSF) output capability are two major criteria characterizing the performance of a power amplifier in the application of multifrequency eddy current testing (MECT). Switch-mode power amplifiers are known to have a very high efficiency, yet they have rarely been adopted in the instrumental development of MECT. In addition, switch-mode power amplifiers themselves are lacking in the research literature for MSF capability. In this article, a Class D power amplifier is designed so as to address the two issues. An MSF selective harmonic elimination pulsewidth modulation method is proposed to generate alternating magnetic fields, which are rich in selected harmonics. A field-programmable-gate-array-based experimental system has been developed to verify the design. Results show that the proposed methodology is capable of generating high MSF currents in the transmitting coil with a low distortion of signal.

eess.SY↗

arXiv2019

Optical Transient Object Classification in Wide Field Small Aperture Telescopes with Neural Networks

Wide field small aperture telescopes are working horses for fast sky surveying. Transient discovery is one of their main tasks. Classification of candidate transient images between real sources and artifacts with high accuracy is an important step for transient discovery. In this paper, we propose two transient classification methods based on neural networks. The first method uses the convolutional neural network without pooling layers to classify transient images with low sampling rate. The second method assumes transient images as one dimensional signals and is based on recurrent neural networks with long short term memory and leaky ReLu activation function in each detection layer. Testing with real observation data, we find that although these two methods can both achieve more than 94% classification accuracy, they have different classification properties for different targets. Based on this result, we propose to use the ensemble learning method to further increase the classification accuracy to more than 97%.

astro-ph.IM↗

arXiv2017

Maximally Frobenius-destabilized vector bundles over smooth algebraic curves

Vector bundles in positive characteristics have a tendency to be destabilized after pulling back by the Frobenius morphism. In this paper, we closely examine vector bundles over curves that are, in an appropriate sense, maximally destabilized by the Frobenius morphism. Then we prove that such bundles of rank 2 exist over any curve in characteristic 3, and are unique up to twisting by a line bundle. We also give an application of such bundles to the study of ample vector bundles, which is valid in all characteristics.

math.AG↗

arXiv2017

Quantum parameters of the geometric Langlands theory

Fix a smooth, complete algebraic curve $X$ over an algebraically closed field $k$ of characteristic zero. To a reductive group $G$ over $k$, we associate an algebraic stack $\operatorname{Par}_G$ of quantum parameters for the geometric Langlands theory. Then we construct a family of (quasi-)twistings parametrized by $\operatorname{Par}_G$, whose module categories give rise to twisted $\mathcal D$-modules on $\operatorname{Bun}_G$ as well as quasi-coherent sheaves on the DG stack $\operatorname{LocSys}_G$.

math.AG↗

arXiv2015

Lifting representations of finite reductive groups: a character relation

Given a connected reductive group $\tilde{G}$ over a finite field $k$, and a semisimple $k$-automorphism $\varepsilon$ of $\tilde{G}$ of finite order, let $G$ denote the connected part of the group of $\varepsilon$-fixed points. Then there exists a lifting from packets of representations of $G(k)$ to packets for $\tilde{G}(k)$. In the case of Deligne-Lusztig representations, we show that this lifting satisfies a character relation analogous to that of Shintani.

math.RT↗

arXiv2014

How to Upgrade Wireless Networks: Small Cells or Massive MIMO?

Radio network deployment and coverage optimization are critical to next-generation wireless networks. In this paper, the problem of optimally deciding on whether to install additional small cells or to upgrade current macrocell base stations (BSs) with massive antenna arrays is studied. This integrated deployment problem is cast as a general integer optimization model by using the facility location framework. The capacity limits of both the radio access link and the backhaul link are considered. The problem is shown to be an extension of the modular capacitated location problem (MCLP) which is known to be NP-hard. To solve the problem, a novel deployment algorithm that uses Lagrangian relaxation and tabu local search is proposed. The developed tabu search is shown to have a two-level structure and to be able to search the solution space thoroughly. Simulation results show how the proposed, optimal approach to upgrading an existing wireless network infrastructure can make use of a combination of both small cells and BSs with massive antennas. The results also show that the proposed algorithm can find the optimal solution effectively while having a computational time that is up to 30% lower than that of conventional algorithms.

cs.IT↗

arXiv2014

Limiting behavior of Donaldson's heat flow on non-Kähler surfaces

Let $X$ be a compact Hermitian surface, and $g$ be any fixed Gauduchon metric on $X$. Let $E$ be an Hermitian holomorphic vector bundle over $X$. On the bundle $E$, Donaldson's heat flow is gauge equivalent to a flow of holomorphic structures. We prove that this flow converges, in the sense of Uhlenbeck, to the double dual of the graded sheaf associated to the $g$-Harder-Narasimhan-Seshadri filtration of $X$. This result generalizes a convergence theorem of Daskalopoulos and Wentworth to non-Kähler setting.

math.DG↗