arXiv subjects

Yiwei Zhang

Publications and source records attributed to Yiwei Zhang.

At least 91 records · Page 5Linked to original sources

arXiv2021

Room to Grow: Understanding Personal Characteristics Behind Self Improvement Using Social Media

Many people aim for change, but not everyone succeeds. While there are a number of social psychology theories that propose motivation-related characteristics of those who persist with change, few computational studies have explored the motivational stage of personal change. In this paper, we investigate a new dataset consisting of the writings of people who manifest intention to change, some of whom persist while others do not. Using a variety of linguistic analysis techniques, we first examine the writing patterns that distinguish the two groups of people. Persistent people tend to reference more topics related to long-term self-improvement and use a more complicated writing style. Drawing on these consistent differences, we build a classifier that can reliably identify the people more likely to persist, based on their language. Our experiments provide new insights into the motivation-related behavior of people who persist with their intention to change.

cs.CL↗

arXiv2020

Precise asymptotics on the Birkhoff sums for dynamical systems

We establish two precise asymptotic results on the Birkhoff sums for dynamical systems. These results are parallel to that on the arithmetic sums of independent and identically distributed random variables previously obtained by Hsu and Robbins, Erdős, Heyde. We apply our results to the Gauss map and obtain new precise asymptotics in the theorem of Lévy on the regular continued fraction expansion of irrational numbers in $(0,1)$.

math.DS↗

arXiv2020

Look, Listen, and Act: Towards Audio-Visual Embodied Navigation

A crucial ability of mobile intelligent agents is to integrate the evidence from multiple sensory inputs in an environment and to make a sequence of actions to reach their goals. In this paper, we attempt to approach the problem of Audio-Visual Embodied Navigation, the task of planning the shortest path from a random starting location in a scene to the sound source in an indoor environment, given only raw egocentric visual and audio sensory data. To accomplish this task, the agent is required to learn from various modalities, i.e. relating the audio signal to the visual environment. Here we describe an approach to audio-visual embodied navigation that takes advantage of both visual and audio pieces of evidence. Our solution is based on three key ideas: a visual perception mapper module that constructs its spatial memory of the environment, a sound perception module that infers the relative location of the sound source from the agent, and a dynamic path planner that plans a sequence of actions based on the audio-visual observations and the spatial memory of the environment to navigate toward the goal. Experimental results on a newly collected Visual-Audio-Room dataset using the simulated multi-modal environment demonstrate the effectiveness of our approach over several competitive baselines.

cs.CV↗

arXiv2020

On Coupling Lemma and Stochastic Properties with Unbounded Observables for 1-d Expanding Maps

In this paper, we establish a coupling lemma for standard families in the setting of piecewise expanding interval maps with countably many branches. Our method merely requires that the expanding map satisfies Chernov's one-step expansion at $q$-scale and eventually covers a magnet interval. Therefore, our approach is particularly powerful for maps whose inverse Jacobian has low regularity and those who does not satisfy the big image property. The main ingredients of our coupling method are two crucial lemmas: the growth lemma in terms of the characteristic $\cZ$ function and the covering ratio lemma over the magnet interval. We first prove the existence of an absolutely continuous invariant measure. What is more important, we further show that the growth lemma enables the liftablity of the Lebesgue measure to the associated Hofbauer tower, and the resulting invariant measure on the tower admits a decomposition of Pesin-Sinai type. Furthermore, we obtain the exponential decay of correlations and the almost sure invariance principle (which is a functional version of the central limit theorem). For the first time, we are able to make a direct relation between the mixing rates and the $\cZ$ function, see (\ref{equ:totalvariation1}). The novelty of our results relies on establishing the regularity of invariant density, as well as verifying the stochastic properties for a large class of unbounded observables. Finally, we verify our assumptions for several well known examples that were previously studied in the literature, and unify results to these examples in our framework.

math.DS↗

arXiv2020

Measuring Similarity between Brands using Followers' Post in Social Media

In this paper, we propose a new measure to estimate the similarity between brands via posts of brands' followers on social network services (SNS). Our method was developed with the intention of exploring the brands that customers are likely to jointly purchase. Nowadays, brands use social media for targeted advertising because influencing users' preferences can greatly affect the trends in sales. We assume that data on SNS allows us to make quantitative comparisons between brands. Our proposed algorithm analyzes the daily photos and hashtags posted by each brand's followers. By clustering them and converting them to histograms, we can calculate the similarity between brands. We evaluated our proposed algorithm with purchase logs, credit card information, and answers to the questionnaires. The experimental results show that the purchase data maintained by a mall or a credit card company can predict the co-purchase very well, but not the customer's willingness to buy products of new brands. On the other hand, our method can predict the users' interest on brands with a correlation value over 0.53, which is pretty high considering that such interest to brands are high subjective and individual dependent.

cs.SI↗

arXiv2019

Factorized Multimodal Transformer for Multimodal Sequential Learning

The complex world around us is inherently multimodal and sequential (continuous). Information is scattered across different modalities and requires multiple continuous sensors to be captured. As machine learning leaps towards better generalization to real world, multimodal sequential learning becomes a fundamental research area. Arguably, modeling arbitrarily distributed spatio-temporal dynamics within and across modalities is the biggest challenge in this research area. In this paper, we present a new transformer model, called the Factorized Multimodal Transformer (FMT) for multimodal sequential learning. FMT inherently models the intramodal and intermodal (involving two or more modalities) dynamics within its multimodal input in a factorized manner. The proposed factorization allows for increasing the number of self-attentions to better model the multimodal phenomena at hand; without encountering difficulties during training (e.g. overfitting) even on relatively low-resource setups. All the attention mechanisms within FMT have a full time-domain receptive field which allows them to asynchronously capture long-range multimodal dynamics. In our experiments we focus on datasets that contain the three commonly studied modalities of language, vision and acoustic. We perform a wide range of experiments, spanning across 3 well-studied datasets and 21 distinct labels. FMT shows superior performance over previously proposed models, setting new state of the art in the studied datasets.

cs.LG↗

arXiv2019

User-Aware Folk Popularity Rank: User-Popularity-Based Tag Recommendation That Can Enhance Social Popularity

In this paper we propose a method that can enhance the social popularity of a post (i.e., the number of views or likes) by recommending appropriate hash tags considering both content popularity and user popularity. A previous approach called FolkPopularityRank (FP-Rank) considered only the relationship among images, tags, and their popularity. However, the popularity of an image/video is strongly affected by who uploaded it. Therefore, we develop an algorithm that can incorporate user popularity and users' tag usage tendency into the FP-Rank algorithm. The experimental results using 60,000 training images with their accompanying tags and 1,000 test data, which were actually uploaded to a real social network service (SNS), show that, in ten days, our proposed algorithm can achieve 1.2 times more views than the FP-Rank algorithm. This technology would be critical to individual users and companies/brands who want to promote themselves in SNSs.

cs.SI↗

arXiv2019

Ergodic optimization theory for Axiom A flows

In this article, we consider the weighted ergodic optimization problem Axiom A attractors of a $C^2$ flow on a compact smooth manifold. The main result obtained in this paper is that for a generic observable from function space $\mc C^{0,\a}$ ($\a\in(0,1]$) or $\mc C^1$ the minimizing measure is unique and is supported on a periodic orbit.

math.DS↗

arXiv2019

Ergodic optimization theory for a class of typical maps

In this article, we consider the weighted ergodic optimization problem of a class of dynamical systems $T:X\to X$ where $X$ is a compact metric space and $T$ is Lipschitz continuous. We show that once $T:X\to X$ satisfies both the {\em Anosov shadowing property }({\bf ASP}) and the {\em Mañé-Conze-Guivarc'h-Bousch property }({\bf MCGBP}), the minimizing measures of generic Hölder observables are unique and supported on a periodic orbit. Moreover, if $T:X\to X$ is a subsystem of a dynamical system $f:M\to M$ (i.e. $X\subset M$ and $f|_X=T$) where $M$ is a compact smooth manifold, the above conclusion holds for $C^1$ observables. Note that a broad class of classical dynamical systems satisfies both ASP and MCGBP, which includes {\em Axiom A attractors, Anosov diffeomorphisms }and {\em uniformly expanding maps}. Therefore, the open problem proposed by Yuan and Hunt in \cite{YH} for $C^1$-observables is solved consequentially.

math.DS↗

arXiv2019

Private Proximity Retrieval Codes

A \emph{private proximity retrieval} (\emph{PPR}) scheme is a protocol which allows a user to retrieve the identities of all records in a database that are within some distance $r$ from the user's record $x$. The user's \emph{privacy} at each server is given by the fraction of the record $x$ that is kept private. In this paper, this research is initiated and protocols that offer trade-offs between privacy and computational complexity and storage are studied. In particular, we assume that each server stores a copy of the database and study the required minimum number of servers by our protocol which provides a given privacy level. Each server gets a query in the protocol and the set of queries forms a code. We study the family of codes generated by the set of queries and in particular the minimum number of codewords in such a code which is the minimum number of servers required for the protocol. These codes are closely related to a family of codes known as \emph{covering designs}. We introduce several lower bounds on the sizes of such codes as well as several constructions. This work focuses on the case when the records are binary vectors together with the Hamming distance. Other metrics such as the Johnson metric are also investigated.

cs.IT↗

arXiv2019

Anomalous time-scaling of extreme events in infinite systems and Birkhoff sums of infinite observables

We establish quantitative results for the statistical be\-ha\-vi\-our of \emph{infinite systems}. We consider two kinds of infinite system: i) a conservative dynamical system $(f,X,μ)$ preserving a $σ$-finite measure $μ$ such that $μ(X)=\infty$; ii) the case where $μ$ is a probability measure but we consider the statistical behaviour of an observable $ϕ\colon X\to[0,\infty)$ which is non-integrable: $\int ϕ\, dμ=\infty$. In the first part of this work we study the behaviour of Birkhoff sums of systems of the kind ii). For certain weakly chaotic systems, we show that these sums can be strongly oscillating. However, if the system has superpolynomial decay of correlations or has a Markov structure, then we show this oscillation cannot happen. In this case we prove asymptotic relations between the behaviour of $ϕ$, the local dimension of $μ$, and on the growth of Birkhoff sums (as time tends to infinity). We then establish several important consequences which apply to infinite systems of the kind i). This includes showing anomalous scalings in extreme event limit laws, or entrance time statistics. We apply our findings to non-uniformly hyperbolic systems preserving an infinite measure, establishing anomalous scalings in the case of logarithm laws of entrance times, dynamical Borel--Cantelli lemmas, almost sure growth rates of extremes, and dynamical run length functions.

math.DS↗

arXiv2019

Bounds on the Length of Functional PIR and Batch codes

A functional $k$-PIR code of dimension $s$ consists of $n$ servers storing linear combinations of $s$ linearly independent information symbols. Any linear combination of the $s$ information symbols can be recovered by $k$ disjoint subsets of servers. The goal is to find the smallest number of servers for given $k$ and $s$. We provide lower bounds on the number of servers and constructions which yield upper bounds on this number. For $k \leq 4$, exact bounds on the number of servers are proved. Furthermore, we provide some asymptotic bounds. The problem coincides with the well known private information retrieval problem based on a coded database to reduce the storage overhead, when each linear combination contains exactly one information symbol. If any multiset of size $k$ of linear combinations from the linearly independent information symbols can be recovered by $k$ disjoint subset of servers, then the servers form a functional $k$-batch code. A~functional $k$-batch code is a functional $k$-PIR code, where all the $k$ linear combinations in the multiset are equal. We provide some bounds on the number of servers for functional $k$-batch codes. In particular we present a random construction and a construction based on simplex codes, WOM codes, and RIO codes.

cs.IT↗

arXiv2019

Spectral gap of sparse bistochastic matrices with exchangeable rows with application to shuffle-and-fold maps

We consider a random bistochastic matrix of size $n$ of the form $M Q$ where $M$ is a uniformly distributed permutation matrix and $Q$ is a given bistochastic matrix. Under mild sparsity and regularity assumptions on $Q$, we prove that the second largest eigenvalue of $MQ$ is essentially bounded by the normalized Hilbert-Schmidt norm of $Q$ when $n$ grows large. We apply this result to random walks on random regular digraphs and to shuffle-and-fold maps of the unit interval popularized in fluid mixing protocols.

math.DS↗

arXiv2019

Bi-Skip: A Motion Deblurring Network Using Self-paced Learning

A fast and effective motion deblurring method has great application values in real life. This work presents an innovative approach in which a self-paced learning is combined with GAN to deblur image. First, We explain that a proper generator can be used as deep priors and point out that the solution for pixel-based loss is not same with the one for perception-based loss. By using these ideas as starting points, a Bi-Skip network is proposed to improve the generating ability and a bi-level loss is adopted to solve the problem that common conditions are non-identical. Second, considering that the complex motion blur will perturb the network in the training process, a self-paced mechanism is adopted to enhance the robustness of the network. Through extensive evaluations on both qualitative and quantitative criteria, it is demonstrated that our approach has a competitive advantage over state-of-the-art methods.

cs.CV↗

arXiv2019

On the Access Complexity of PIR Schemes

Private information retrieval has been reformulated in an information-theoretic perspective in recent years. The two most important parameters considered for a PIR scheme in a distributed storage system are the storage overhead and PIR rate. The complexity of the computations done by the servers for the various tasks of the distributed storage system is an important parameter in such systems which didn't get enough attention in PIR schemes. As a consequence, we take into consideration a third parameter, the access complexity of a PIR scheme, which characterizes the total amount of data to be accessed by the servers for responding to the queries throughout a PIR scheme. We use a general covering codes approach as the main tool for improving the access complexity. With a given amount of storage overhead, the ultimate objective is to characterize the tradeoff between the rate and access complexity of a PIR scheme. This covering codes approach raises a new interesting coding problem of generalized coverings similarly to the well-known generalized Hamming weights.

cs.IT↗

arXiv2018

Detection based Defense against Adversarial Examples from the Steganalysis Point of View

Deep Neural Networks (DNNs) have recently led to significant improvements in many fields. However, DNNs are vulnerable to adversarial examples which are samples with imperceptible perturbations while dramatically misleading the DNNs. Moreover, adversarial examples can be used to perform an attack on various kinds of DNN based systems, even if the adversary has no access to the underlying model. Many defense methods have been proposed, such as obfuscating gradients of the networks or detecting adversarial examples. However it is proved out that these defense methods are not effective or cannot resist secondary adversarial attacks. In this paper, we point out that steganalysis can be applied to adversarial examples detection, and propose a method to enhance steganalysis features by estimating the probability of modifications caused by adversarial attacks. Experimental results show that the proposed method can accurately detect adversarial examples. Moreover, secondary adversarial attacks cannot be directly performed to our method because our method is not based on a neural network but based on high-dimensional artificial features and FLD (Fisher Linear Discriminant) ensemble.

cs.CV↗

arXiv2018

A note on the run length function for intermittency maps

We study the run length function for intermittency maps. In particular, we show that the longest consecutive zero digits (resp. one digits) having a time window of polynomial (resp. logarithmic) length. Our proof is relatively elementary in the sense that it only relies on the classical Borel-Cantelli lemma and the polynomial decay of intermittency maps. Our results are compensational to the Erdős-Rényi law obtained by Denker and Nicol in \cite{dennic13}.

math.DS↗

arXiv2018

New Theoretical Bounds and Constructions of Permutation Codes under Block Permutation Metric

Permutation codes under different metrics have been extensively studied due to their potentials in various applications. Generalized Cayley metric is introduced to correct generalized transposition errors, including previously studied metrics such as Kendall's $τ$-metric, Ulam metric and Cayley metric as special cases. Since the generalized Cayley distance between two permutations is not easily computable, Yang et al. introduced a related metric of the same order, named the block permutation metric. Given positive integers $n$ and $d$, let $\mathcal{C}_{B}(n,d)$ denote the maximum size of a permutation code in $S_n$ with minimum block permutation distance $d$. In this paper, we focus on the theoretical bounds of $\mathcal{C}_{B}(n,d)$ and the constructions of permutation codes under block permutation metric. Using a graph theoretic approach, we improve the Gilbert-Varshamov type bound by a factor of $Ω(\log{n})$, when $d$ is fixed and $n$ goes into infinity. We also propose a new encoding scheme based on binary constant weight codes. Moreover, an upper bound beating the sphere-packing type bound is given when $d$ is relatively close to $n$.

cs.IT↗