Search arXiv⌕ Search

arXiv subjects

Ling Chen

Publications and source records attributed to Ling Chen.

198 records · Page 11Linked to original sources
arXiv2018

Privacy-preserving Stochastic Gradual Learning

It is challenging for stochastic optimizations to handle large-scale sensitive data safely. Recently, Duchi et al. proposed private sampling strategy to solve privacy leakage in stochastic optimizations. However, this strategy leads to robustness degeneration, since this strategy is equal to the noise injection on each gradient, which adversely affects updates of the primal variable. To address this challenge, we introduce a robust stochastic optimization under the framework of local privacy, which is called Privacy-pREserving StochasTIc Gradual lEarning (PRESTIGE). PRESTIGE bridges private updates of the primal variable (by private sampling) with the gradual curriculum learning (CL). Specifically, the noise injection leads to the issue of label noise, but the robust learning process of CL can combat with label noise. Thus, PRESTIGE yields "private but robust" updates of the primal variable on the private curriculum, namely an reordered label sequence provided by CL. In theory, we reveal the convergence rate and maximum complexity of PRESTIGE. Empirical results on six datasets show that, PRESTIGE achieves a good tradeoff between privacy preservation and robustness over baselines.

cs.CR↗
arXiv2018

Learning Representations of Ultrahigh-dimensional Data for Random Distance-based Outlier Detection

Learning expressive low-dimensional representations of ultrahigh-dimensional data, e.g., data with thousands/millions of features, has been a major way to enable learning methods to address the curse of dimensionality. However, existing unsupervised representation learning methods mainly focus on preserving the data regularity information and learning the representations independently of subsequent outlier detection methods, which can result in suboptimal and unstable performance of detecting irregularities (i.e., outliers). This paper introduces a ranking model-based framework, called RAMODO, to address this issue. RAMODO unifies representation learning and outlier detection to learn low-dimensional representations that are tailored for a state-of-the-art outlier detection approach - the random distance-based approach. This customized learning yields more optimal and stable representations for the targeted outlier detectors. Additionally, RAMODO can leverage little labeled data as prior knowledge to learn more expressive and application-relevant representations. We instantiate RAMODO to an efficient method called REPEN to demonstrate the performance of RAMODO. Extensive empirical results on eight real-world ultrahigh dimensional data sets show that REPEN (i) enables a random distance-based detector to obtain significantly better AUC performance and two orders of magnitude speedup; (ii) performs substantially better and more stably than four state-of-the-art representation learning methods; and (iii) leverages less than 1% labeled data to achieve up to 32% AUC improvement.

cs.LG↗
arXiv2017

Improved Consistent Weighted Sampling Revisited

Min-Hash is a popular technique for efficiently estimating the Jaccard similarity of binary sets. Consistent Weighted Sampling (CWS) generalizes the Min-Hash scheme to sketch weighted sets and has drawn increasing interest from the community. Due to its constant-time complexity independent of the values of the weights, Improved CWS (ICWS) is considered as the state-of-the-art CWS algorithm. In this paper, we revisit ICWS and analyze its underlying mechanism to show that there actually exists dependence between the two components of the hash-code produced by ICWS, which violates the condition of independence. To remedy the problem, we propose an Improved ICWS (I$^2$CWS) algorithm which not only shares the same theoretical computational complexity as ICWS but also abides by the required conditions of the CWS scheme. The experimental results on a number of synthetic data sets and real-world text data sets demonstrate that our I$^2$CWS algorithm can estimate the Jaccard similarity more accurately, and also compete with or outperform the compared methods, including ICWS, in classification and top-$K$ retrieval, after relieving the underlying dependence.

cs.DS↗
arXiv2017

Structured Peer Learning Program - An Innovative Approach to Computer Science Education

Structured Peer Learning (SPL) is a form of peer-based supplemental instruction that focuses on mentoring, guidance, and development of technical, communication, and social skills in both the students receiving assistance and the students in teaching roles. This paper explores the methodology, efficacy, and reasoning behind the practical realization of a SPL program designed to increase student knowledge and success in undergraduate Computer Science courses. Students expressed an increased level of comfort when asking for help from student teachers versus traditional educational resources, historically showed an increased average grade in lower-level courses, and felt that the program positively impacted their desire to continue in or switch to a Computer major. Additionally, results indicated that advances in programming, analytical thinking, and abstract analysis skills were evident in not only the students but also the student teachers, suggesting a strong bidirectional flow of knowledge.

cs.CY↗
arXiv2016

Extracting Actionability from Machine Learning Models by Sub-optimal Deterministic Planning

A main focus of machine learning research has been improving the generalization accuracy and efficiency of prediction models. Many models such as SVM, random forest, and deep neural nets have been proposed and achieved great success. However, what emerges as missing in many applications is actionability, i.e., the ability to turn prediction results into actions. For example, in applications such as customer relationship management, clinical prediction, and advertisement, the users need not only accurate prediction, but also actionable instructions which can transfer an input to a desirable goal (e.g., higher profit repays, lower morbidity rates, higher ads hit rates). Existing effort in deriving such actionable knowledge is few and limited to simple action models which restricted to only change one attribute for each action. The dilemma is that in many real applications those action models are often more complex and harder to extract an optimal solution. In this paper, we propose a novel approach that achieves actionability by combining learning with planning, two core areas of AI. In particular, we propose a framework to extract actionable knowledge from random forest, one of the most widely used and best off-the-shelf classifiers. We formulate the actionability problem to a sub-optimal action planning (SOAP) problem, which is to find a plan to alter certain features of a given input so that the random forest would yield a desirable output, while minimizing the total costs of actions. Technically, the SOAP problem is formulated in the SAS+ planning formalism, and solved using a Max-SAT based approach. Our experimental results demonstrate the effectiveness and efficiency of the proposed approach on a personal credit dataset and other benchmarks. Our work represents a new application of automated planning on an emerging and challenging machine learning paradigm.

cs.AI↗
arXiv2016

On the Convergence of A Family of Robust Losses for Stochastic Gradient Descent

The convergence of Stochastic Gradient Descent (SGD) using convex loss functions has been widely studied. However, vanilla SGD methods using convex losses cannot perform well with noisy labels, which adversely affect the update of the primal variable in SGD methods. Unfortunately, noisy labels are ubiquitous in real world applications such as crowdsourcing. To handle noisy labels, in this paper, we present a family of robust losses for SGD methods. By employing our robust losses, SGD methods successfully reduce negative effects caused by noisy labels on each update of the primal variable. We not only reveal that the convergence rate is O(1/T) for SGD methods using robust losses, but also provide the robustness analysis on two representative robust losses. Comprehensive experimental results on six real-world datasets show that SGD methods using robust losses are obviously more robust than other baseline methods in most situations with fast convergence.

cs.LG↗
arXiv2015

WaveCluster with Differential Privacy

WaveCluster is an important family of grid-based clustering algorithms that are capable of finding clusters of arbitrary shapes. In this paper, we investigate techniques to perform WaveCluster while ensuring differential privacy. Our goal is to develop a general technique for achieving differential privacy on WaveCluster that accommodates different wavelet transforms. We show that straightforward techniques based on synthetic data generation and introduction of random noise when quantizing the data, though generally preserving the distribution of data, often introduce too much noise to preserve useful clusters. We then propose two optimized techniques, PrivTHR and PrivTHREM, which can significantly reduce data distortion during two key steps of WaveCluster: the quantization step and the significant grid identification step. We conduct extensive experiments based on four datasets that are particularly interesting in the context of clustering, and show that PrivTHR and PrivTHREM achieve high utility when privacy budgets are properly allocated.

cs.DB↗
arXiv2015

An $S_3$-symmetry of the Jacobi Identity for Intertwining Operator Algebras

We prove an $S_{3}$-symmetry of the Jacobi identity for intertwining operator algebras. Since this Jacobi identity involves the braiding and fusing isomorphisms satisfying the genus-zero Moore-Seiberg equations, our proof uses not only the basic properties of intertwining operators, but also the properties of braiding and fusing isomorphisms and the genus-zero Moore-Seiberg equations. Our proof depends heavily on the theory of multivalued analytic functions of several variables, especially the theory of analytic extensions.

math.QA↗
arXiv2015

On Axiomatic Approaches to Intertwining Operator Algebras

We study intertwining operator algebras introduced and constructed by Huang. In the case that the intertwining operator algebras involve intertwining operators among irreducible modules for their vertex operator subalgebras, a number of results on intertwining operator algebras were given in [H9] but some of the proofs were postponed to an unpublished monograph. In this paper, we give the proofs of these results in [H9] and we formulate and prove results for general intertwining operator algebras without assuming that the modules involved are irreducible. In particular, we construct fusing and braiding isomorphisms for general intertwining operator algebras and prove that they satisfy the genus-zero Moore-Seiberg equations. We show that the Jacobi identity for intertwining operator algebras is equivalent to generalized rationality, commutativity and associativity properties of intertwining operator algebras. We introduce the locality for intertwining operator algebras and show that the Jacobi identity is equivalent to the locality, assuming that other axioms hold. Moreover, we establish that any two of the three properties, associativity, commutativity and skew-symmetry, imply the other (except that when deriving skew-symmetry from associativity and commutativity, more conditions are needed). Finally, we show that three definitions of intertwining operator algebras are equivalent.

math.QA↗
arXiv2015

Geo-SAGE: A Geographical Sparse Additive Generative Model for Spatial Item Recommendation

With the rapid development of location-based social networks (LBSNs), spatial item recommendation has become an important means to help people discover attractive and interesting venues and events, especially when users travel out of town. However, this recommendation is very challenging compared to the traditional recommender systems. A user can visit only a limited number of spatial items, leading to a very sparse user-item matrix. Most of the items visited by a user are located within a short distance from where he/she lives, which makes it hard to recommend items when the user travels to a far away place. Moreover, user interests and behavior patterns may vary dramatically across different geographical regions. In light of this, we propose Geo-SAGE, a geographical sparse additive generative model for spatial item recommendation in this paper. Geo-SAGE considers both user personal interests and the preference of the crowd in the target region, by exploiting both the co-occurrence pattern of spatial items and the content of spatial items. To further alleviate the data sparsity issue, Geo-SAGE exploits the geographical correlation by smoothing the crowd's preferences over a well-designed spatial index structure called spatial pyramid. We conduct extensive experiments to evaluate the performance of our Geo-SAGE model on two real large-scale datasets. The experimental results clearly demonstrate our Geo-SAGE model outperforms the state-of-the-art in the two tasks of both out-of-town and home-town recommendations.

cs.IR↗
arXiv2015

Link Prediction in Networks with Nodes Attributes by Similarity Propagation

The problem of link prediction has attracted considerable recent attention from various domains such as sociology, anthropology, information science, and computer sciences. A link prediction algorithm is proposed based on link similarity score propagation by a random walk in networks with nodes attributes. In the algorithm, each link in the network is assigned a transmission probability according to the similarity of the attributes on the nodes connected by the link. The link similarity score between the nodes are then propagated via the links according to their transmission probability. Our experimental results show that it can obtain higher quality results on the networks with node attributes than other algorithms.

cs.SI↗
arXiv2014

Excitation of Langmuir waves by the lower energy cutoff behavior of power-law electrons

Langmuir waves (LWs), which are believed to play a crucial role in the plasma emission of solar radio bursts, can be excited by streaming instability of energetic electron beams. However, solar hard X-ray observations imply that the energetic flare electrons usually have a power-law energy distribution with a lower energy cutoff. In this paper, we investigate LWs driven by the power-law electrons. The results show that power-law electrons with the steepness cutoff behavior can excite LWs effectively because of the population inversion distribution below the cutoff energy ($E_c$). The growth rate of LWs increases with the steepness index ($δ$) and decreases with the power-law index ($α$). The wave number of the fastest growing LWs ($kλ_D$), decreases with the characteristic velocity of the power-law electrons ($v_{c}=\sqrt{2E_{c}/m_{e}}$) and increases with the thermal velocity of ambient electrons ($v_T$). This can be helpful for us to understand better the physics of LWs and the dynamics of energetic electron beams in space and astrophysical plasmas.

astro-ph.SR↗
arXiv2012

Some Representations of Nongraded Divergence-Free Lie Algebras

Divergence-free Lie algebras are originated from the Lie algebras of volume-preserving transformation groups. Xu constructed a certain nongraded generalization, which may not contain any toral Cartan subalgebra. In this paper, we give a complete classification of the generalized weight modules over these algebras with weight multiplicities less than or equal to one.

math.RT↗
arXiv2010

Multiplicity-Free Representations of Divergence-Free Lie Algebras

Divergence-free Lie algebras (also known as the special Lie algebras of Cartan type) are Lie algebras of volume-preserving transformation groups. They are simple in generic case. Dokovic and Zhao found a certain graded generalization of them. In this paper, we classify all the irreducible and indecomposable multiplicity-free modules of the simple generalized divergence-free Lie algebras.

math.QA↗
arXiv2010

Twisted Hamiltonian Lie Algebras and Their Multiplicity-Free Representations

We construct a class of new Lie algebras by generalizing the one-variable Lie algebras generated by the quadratic conformal algebras (or corresponding Hamiltonian operators) associated to Poisson algebras and a quasi-derivation found by Xu. These algebras can be viewed as certain twists of Xu's generalized Hamiltonian Lie algebras. The simplicity of these algebras is completely determined. Moreover, we construct a family of multiplicity-free representations of these Lie algebras and prove their irreducibility.

math.QA↗
arXiv2006

Electronic charge reconstruction of doped Mott insulators in multilayered nanostructures

Dynamical mean-field theory is employed to calculate the electronic charge reconstruction of multilayered inhomogeneous devices composed of semi-infinite metallic lead layers sandwiching barrier planes of a strongly correlated material (that can be tuned through the metal-insulator Mott transition). The main focus is on barriers that are doped Mott insulators, and how the electronic charge reconstruction can create well-defined Mott insulating regions in a device whose thickness is governed by intrinsic materials properties, and hence may be able to be reproducibly made.

cond-mat.str-el↗
arXiv2003

Charge-density-wave order parameter of the Falicov-Kimball model in infinite dimensions

In the large-U limit, the Falicov-Kimball model maps onto an effective Ising model, with an order parameter described by a BCS-like mean-field theory in infinite dimensions. In the small-U limit, van Dongen and Vollhardt showed that the order parameter assumes a strange non-BCS-like shape with a sharp reduction near T approx T_c/2. Here we numerically investigate the crossover between these two regimes and qualitatively determine the order parameter for a variety of different values of U. We find the overall behavior of the order parameter as a function of temperature to be quite anomalous.

cond-mat.str-el↗