arXiv subjects

Bo Tang

Publications and source records attributed to Bo Tang.

At least 145 records · Page 8Linked to original sources

arXiv2021

Stragglers Are Not Disaster: A Hybrid Federated Learning Algorithm with Delayed Gradients

Federated learning (FL) is a new machine learning framework which trains a joint model across a large amount of decentralized computing devices. Existing methods, e.g., Federated Averaging (FedAvg), are able to provide an optimization guarantee by synchronously training the joint model, but usually suffer from stragglers, i.e., IoT devices with low computing power or communication bandwidth, especially on heterogeneous optimization problems. To mitigate the influence of stragglers, this paper presents a novel FL algorithm, namely Hybrid Federated Learning (HFL), to achieve a learning balance in efficiency and effectiveness. It consists of two major components: synchronous kernel and asynchronous updater. Unlike traditional synchronous FL methods, our HFL introduces the asynchronous updater which actively pulls unsynchronized and delayed local weights from stragglers. An adaptive approximation method, Adaptive Delayed-SGD (AD-SGD), is proposed to merge the delayed local updates into the joint model. The theoretical analysis of HFL shows that the convergence rate of the proposed algorithm is $\mathcal{O}(\frac{1}{t+τ})$ for both convex and non-convex optimization problems.

cs.LG↗

arXiv2021

Overcoming Long-term Catastrophic Forgetting through Adversarial Neural Pruning and Synaptic Consolidation

Artificial neural networks face the well-known problem of catastrophic forgetting. What's worse, the degradation of previously learned skills becomes more severe as the task sequence increases, known as the long-term catastrophic forgetting. It is due to two facts: first, as the model learns more tasks, the intersection of the low-error parameter subspace satisfying for these tasks becomes smaller or even does not exist; second, when the model learns a new task, the cumulative error keeps increasing as the model tries to protect the parameter configuration of previous tasks from interference. Inspired by the memory consolidation mechanism in mammalian brains with synaptic plasticity, we propose a confrontation mechanism in which Adversarial Neural Pruning and synaptic Consolidation (ANPyC) is used to overcome the long-term catastrophic forgetting issue. The neural pruning acts as long-term depression to prune task-irrelevant parameters, while the novel synaptic consolidation acts as long-term potentiation to strengthen task-relevant parameters. During the training, this confrontation achieves a balance in that only crucial parameters remain, and non-significant parameters are freed to learn subsequent tasks. ANPyC avoids forgetting important information and makes the model efficient to learn a large number of tasks. Specifically, the neural pruning iteratively relaxes the current task's parameter conditions to expand the common parameter subspace of the task; the synaptic consolidation strategy, which consists of a structure-aware parameter-importance measurement and an element-wise parameter updating strategy, decreases the cumulative error when learning new tasks. The full source code is available at https://github.com/GeoX-Lab/ANPyC.

cs.LG↗

arXiv2020

Towards Self-Tuning Parameter Servers

Recent years, many applications have been driven advances by the use of Machine Learning (ML). Nowadays, it is common to see industrial-strength machine learning jobs that involve millions of model parameters, terabytes of training data, and weeks of training. Good efficiency, i.e., fast completion time of running a specific ML job, therefore, is a key feature of a successful ML system. While the completion time of a long-running ML job is determined by the time required to reach model convergence, practically that is also largely influenced by the values of various system settings. In this paper, we contribute techniques towards building self-tuning parameter servers. Parameter Server (PS) is a popular system architecture for large-scale machine learning systems; and by self-tuning we mean while a long-running ML job is iteratively training the expert-suggested model, the system is also iteratively learning which system setting is more efficient for that job and applies it online. While our techniques are general enough to various PS-style ML systems, we have prototyped our techniques on top of TensorFlow. Experiments show that our techniques can reduce the completion times of a variety of long-running TensorFlow jobs from 1.4x to 18x.

cs.DB↗

arXiv2020

Learning Discontinuous Piecewise Affine Fitting Functions using Mixed Integer Programming for Segmentation and Denoising

Piecewise affine functions are widely used to approximate nonlinear and discontinuous functions. However, most, if not all existing models only deal with fitting continuous functions. In this paper, we investigate the problem of fitting a discontinuous piecewise affine function to given data that lie in an orthogonal grid, where no restriction on the partition is enforced (i.e., its geometric shape can be nonconvex). This is useful for segmentation and denoising when data corresponding to images. We propose a novel Mixed Integer Program (MIP) formulation for the piecewise affine fitting problem, where binary variables determine the location of break-points. To obtain consistent partitions (i.e. image segmentation), we include multi-cut constraints in the formulation. Since the resulting problem is $\mathcal{NP}$-hard, two techniques are introduced to improve the computation. One is to add facet-defining inequalities to the formulation and the other to provide initial integer solutions using a special heuristic algorithm. We conduct extensive experiments by some synthetic images as well as real depth images, and the results demonstrate the feasibility of our model.

math.OC↗

arXiv2020

RCELF: A Residual-based Approach for Influence Maximization Problem

Influence Maximization Problem (IMP) is selecting a seed set of nodes in the social network to spread the influence as widely as possible. It has many applications in multiple domains, e.g., viral marketing is frequently used for new products or activities advertisements. While it is a classic and well-studied problem in computer science, unfortunately, all those proposed techniques are compromising among time efficiency, memory consumption, and result quality. In this paper, we conduct comprehensive experimental studies on the state-of-the-art IMP approximate approaches to reveal the underlying trade-off strategies. Interestingly, we find that even the state-of-the-art approaches are impractical when the propagation probability of the network have been taken into consideration. With the findings of existing approaches, we propose a novel residual-based approach (i.e., RCELF) for IMP, which i) overcomes the deficiencies of existing approximate approaches, and ii) provides theoretical guaranteed results with high efficiency in both time- and space- perspectives. We demonstrate the superiority of our proposal by extensive experimental evaluation on real datasets.

cs.DB↗

arXiv2020

Polyphase Waveform Design for MIMO Radar Space Time Adaptive Processing

We consider the design of polyphase waveforms for ground moving target detection with airborne multiple-input-multiple-output (MIMO) radar. Due to the constant-modulus and finite-alphabet constraint on the waveforms, the associated design problem is non-convex and in general NP-hard. To tackle this problem, we develop an efficient algorithm based on relaxation and cyclic optimization. Moreover, we exploit a reparameterization trick to avoid the significant computational burden and memory requirement brought about by relaxation. We prove that the objective values during the iterations are guaranteed to converge. Finally, we provide an effective randomization approach to obtain polyphase waveforms from the relaxed solution at convergence. Numerical examples show the effectiveness of the proposed algorithm for designing polyphase waveforms.

eess.SP↗

arXiv2018

Find a Reasonable Ending for Stories: Does Logic Relation Help the Story Cloze Test?

Natural language understanding is a challenging problem that covers a wide range of tasks. While previous methods generally train each task separately, we consider combining the cross-task features to enhance the task performance. In this paper, we incorporate the logic information with the help of the Natural Language Inference (NLI) task to the Story Cloze Test (SCT). Previous work on SCT considered various semantic information, such as sentiment and topic, but lack the logic information between sentences which is an essential element of stories. Thus we propose to extract the logic information during the course of the story to improve the understanding of the whole story. The logic information is modeled with the help of the NLI task. Experimental results prove the strength of the logic information.

cs.CL↗

arXiv2016

Adaptive Signal Detection and Parameter Estimation in Unknown Colored Gaussian Noise

This paper considers the general signal detection and parameter estimation problem in the presence of colored Gaussian noise disturbance. By modeling the disturbance with an autoregressive process, we present three signal detectors with different unknown parameters under the general framework of binary hypothesis testing. The closed form of parameter estimates and the asymptotic distributions of these three tests are also given. Given two examples of frequency modulated signal detection problem and time series moving object detection problem, the simulation results demonstrate the effectiveness of three presented detectors.

physics.data-an↗

arXiv2016

Detection of False Data Injection Attacks in Smart Grid under Colored Gaussian Noise

In this paper, we consider the problems of state estimation and false data injection detection in smart grid when the measurements are corrupted by colored Gaussian noise. By modeling the noise with the autoregressive process, we estimate the state of the power transmission networks and develop a generalized likelihood ratio test (GLRT) detector for the detection of false data injection attacks. We show that the conventional approach with the assumption of Gaussian noise is a special case of the proposed method, and thus the new approach has more applicability. {The proposed detector is also tested on an independent component analysis (ICA) based unobservable false data attack scheme that utilizes similar assumptions of sample observation.} We evaluate the performance of the proposed state estimator and attack detector on the IEEE 30-bus power system with comparison to conventional Gaussian noise based detector. The superior performance of {both observable and unobservable false data attacks} demonstrates the effectiveness of the proposed approach and indicates a wide application on the power signal processing.

eess.SY↗

arXiv2016

Probabilistic Human Mobility Model in Indoor Environment

Understanding human mobility is important for the development of intelligent mobile service robots as it can provide prior knowledge and predictions of human distribution for robot-assisted activities. In this paper, we propose a probabilistic method to model human motion behaviors which is determined by both internal and external factors in an indoor environment. While the internal factors are represented by the individual preferences, aims and interests, the external factors are indicated by the stimulation of the environment. We model the randomness of human macro-level movement, e.g., the probability of visiting a specific place and staying time, under the Bayesian framework, considering the influence of both internal and external variables. We use two case studies in a shopping mall and in a college student dorm building to show the effectiveness of our proposed probabilistic human mobility model. Real surveillance camera data are used to validate the proposed model together with survey data in the case study of student dorm.

cs.HC↗

arXiv2016

A Local Density-Based Approach for Local Outlier Detection

This paper presents a simple but effective density-based outlier detection approach with the local kernel density estimation (KDE). A Relative Density-based Outlier Score (RDOS) is introduced to measure the local outlierness of objects, in which the density distribution at the location of an object is estimated with a local KDE method based on extended nearest neighbors of the object. Instead of using only $k$ nearest neighbors, we further consider reverse nearest neighbors and shared nearest neighbors of an object for density distribution estimation. Some theoretical properties of the proposed RDOS including its expected value and false alarm probability are derived. A comprehensive experimental study on both synthetic and real-life data sets demonstrates that our approach is more effective than state-of-the-art outlier detection methods.

cs.AI↗

arXiv2016

Kernel-based Generative Learning in Distortion Feature Space

This paper presents a novel kernel-based generative classifier which is defined in a distortion subspace using polynomial series expansion, named Kernel-Distortion (KD) classifier. An iterative kernel selection algorithm is developed to steadily improve classification performance by repeatedly removing and adding kernels. The experimental results on character recognition application not only show that the proposed generative classifier performs better than many existing classifiers, but also illustrate that it has different recognition capability compared to the state-of-the-art discriminative classifier - deep belief network. The recognition diversity indicates that a hybrid combination of the proposed generative classifier and the discriminative classifier could further improve the classification performance. Two hybrid combination methods, cascading and stacking, have been implemented to verify the diversity and the improvement of the proposed classifier.

stat.ML↗

arXiv2016

FSMJ: Feature Selection with Maximum Jensen-Shannon Divergence for Text Categorization

In this paper, we present a new wrapper feature selection approach based on Jensen-Shannon (JS) divergence, termed feature selection with maximum JS-divergence (FSMJ), for text categorization. Unlike most existing feature selection approaches, the proposed FSMJ approach is based on real-valued features which provide more information for discrimination than binary-valued features used in conventional approaches. We show that the FSMJ is a greedy approach and the JS-divergence monotonically increases when more features are selected. We conduct several experiments on real-life data sets, compared with the state-of-the-art feature selection approaches for text categorization. The superior performance of the proposed FSMJ approach demonstrates its effectiveness and further indicates its wide potential applications on data mining.

stat.ML↗

arXiv2016

EEF: Exponentially Embedded Families with Class-Specific Features for Classification

In this letter, we present a novel exponentially embedded families (EEF) based classification method, in which the probability density function (PDF) on raw data is estimated from the PDF on features. With the PDF construction, we show that class-specific features can be used in the proposed classification method, instead of a common feature subset for all classes as used in conventional approaches. We apply the proposed EEF classifier for text categorization as a case study and derive an optimal Bayesian classification rule with class-specific feature selection based on the Information Gain (IG) score. The promising performance on real-life data sets demonstrates the effectiveness of the proposed approach and indicates its wide potential applications.

stat.ML↗

arXiv2016

Toward Optimal Feature Selection in Naive Bayes for Text Categorization

Automated feature selection is important for text categorization to reduce the feature size and to speed up the learning process of classifiers. In this paper, we present a novel and efficient feature selection framework based on the Information Theory, which aims to rank the features with their discriminative capacity for classification. We first revisit two information measures: Kullback-Leibler divergence and Jeffreys divergence for binary hypothesis testing, and analyze their asymptotic properties relating to type I and type II errors of a Bayesian classifier. We then introduce a new divergence measure, called Jeffreys-Multi-Hypothesis (JMH) divergence, to measure multi-distribution divergence for multi-class classification. Based on the JMH-divergence, we develop two efficient feature selection methods, termed maximum discrimination ($MD$) and $MD-χ^2$ methods, for text categorization. The promising results of extensive experiments demonstrate the effectiveness of the proposed approaches.

stat.ML↗

arXiv2015

Well-Supported versus Approximate Nash Equilibria: Query Complexity of Large Games

We study the randomized query complexity of approximate Nash equilibria (ANE) in large games. We prove that, for some constant $ε>0$, any randomized oracle algorithm that computes an $ε$-ANE in a binary-action, $n$-player game must make $2^{Ω(n/\log n)}$ payoff queries. For the stronger solution concept of well-supported Nash equilibria (WSNE), Babichenko previously gave an exponential $2^{Ω(n)}$ lower bound for the randomized query complexity of $ε$-WSNE, for some constant $ε>0$; the same lower bound was shown to hold for $ε$-ANE, but only when $ε=O(1/n)$. Our result answers an open problem posed by Hart and Nisan and Babichenko and is very close to the trivial upper bound of $2^n$. Our proof relies on a generic reduction from the problem of finding an $ε$-WSNE to the problem of finding an $ε/(4α)$-ANE, in large games with $α$ actions, which might be of independent interest.

cs.GT↗

arXiv2015

Tight Bounds for the Price of Anarchy of Simultaneous First Price Auctions

We study the Price of Anarchy of simultaneous first-price auctions for buyers with submodular and subadditive valuations. The current best upper bounds for the Bayesian Price of Anarchy of these auctions are e/(e-1) [Syrgkanis and Tardos 2013] and 2 [Feldman et al. 2013], respectively. We provide matching lower bounds for both cases even for the case of full information and for mixed Nash equilibria via an explicit construction. We present an alternative proof of the upper bound of e/(e-1) for first-price auctions with fractionally subadditive valuations which reveals the worst-case price distribution, that is used as a building block for the matching lower bound construction. We generalize our results to a general class of item bidding auctions that we call bid-dependent auctions (including first-price auctions and all-pay auctions) where the winner is always the highest bidder and each bidder's payment depends only on his own bid. Finally, we apply our techniques to discriminatory price multi-unit auctions. We complement the results of [de Keijzer et al. 2013] for the case of subadditive valuations, by providing a matching lower bound of 2. For the case of submodular valuations, we provide a lower bound of 1.109. For the same class of valuations, we were able to reproduce the upper bound of e/(e-1) using our non-smooth approach.

cs.GT↗

arXiv2015

On the Efficiency of All-Pay Mechanisms

We study the inefficiency of mixed equilibria, expressed as the price of anarchy, of all-pay auctions in three different environments: combinatorial, multi-unit and single-item auctions. First, we consider item-bidding combinatorial auctions where m all-pay auctions run in parallel, one for each good. For fractionally subadditive valuations, we strengthen the upper bound from 2 [Syrgkanis and Tardos STOC'13] to 1.82 by proving some structural properties that characterize the mixed Nash equilibria of the game. Next, we design an all-pay mechanism with a randomized allocation rule for the multi- unit auction. We show that, for bidders with submodular valuations, the mechanism admits a unique, 75% efficient, pure Nash equilibrium. The efficiency of this mechanism outperforms all the known bounds on the price of anarchy of mechanisms used for multi-unit auctions. Finally, we analyze single-item all-pay auctions motivated by their connection to contests and show tight bounds on the price of anarchy of social welfare, revenue and maximum bid.

cs.GT↗