Batch-mode Learning Research Articles

Robust online rain removal for surveillance videos with dynamic rains

The current rain removal techniques proposed for surveillance videos mainly assume consistent rains with invariant extents and types, and are implemented in a batch-mode learning manner. Such assumption deviates from the continuously varying insights of practical rains, and the batch mode further makes the techniques infeasible for real long-lasting videos. To alleviate these issues, this study proposes a novel online rain removal approach to represent practical dynamic rains embedded in surveillance videos. Particularly, we model the rain streaks scattered in each video frame as a patch-wise mixture of Gaussians (P-MoG) distribution, and update its parameters frame by frame. Such a P-MoG modeling manner finely reflects the non-i.i.d. dynamic variations of rains along time. In specific, the P-MoG rain model in each frame is regularized by the learned rain knowledge in previous frames, making the online model adaptable to not-identically-distributed rains in each frame while regularized by not-independently-distributed rains in previous frames. The proposed model is formulated as a concise probabilistic MAP model, which can be readily solved by EM algorithm. We further embed an affine transformation operator into the proposed model, making it adaptable to a wider range of videos with camera jitters. The superiority of the proposed method is substantiated by extensive experiments implemented on synthetic and real videos containing static and dynamic rains as compared with state-of-the-arts in both accuracy and efficiency.

Properties and learning algorithms for faulty RBF networks with coexistence of weight and node failures

Although there are many fault tolerant algorithms for neural networks, they usually focus on one kind of weight failure or node failure only. This paper first proposes a unified fault model for describing the concurrent weight and node failure situation, where open weight fault, open node fault, weight noise, and node noise could happen in a network at the same time. Afterwards, we analyze the training set error of radial basis function (RBF) networks under the concurrent weight and node failure situation. Based on the finding, we define an objective function for tolerating the concurrent weight and node failure situation. We then develop two learning algorithms, one for batch mode learning and one for online mode learning. Furthermore, for the online mode learning, we derive the convergent conditions for two cases, fixed learning rate and adaptive learning rate.

Budget Online Learning Algorithm for Least Squares SVM

Batch-mode least squares support vector machine (LSSVM) is often associated with unbounded number of support vectors (SVs'), making it unsuitable for applications involving large-scale streaming data. Limited-scale LSSVM, which allows efficient updating, seems to be a good solution to tackle this issue. In this paper, to train the limited-scale LSSVM dynamically, we present a budget online LSSVM (BOLSSVM) algorithm. Methodologically, by setting a fixed budget for SVs', we are able to update the LSSVM model according to the updated SVs' set dynamically without retraining from scratch. In particular, when a new small chunk of SVs' substitute for the old ones, the proposed algorithm employs a low rank correction technology and the Sherman-Morrison-Woodbury formula to compute the inverse of saddle point matrix derived from the LSSVM's Karush-Kuhn-Tucker (KKT) system, which, in turn, updates the LSSVM model efficiently. In this way, the proposed BOLSSVM algorithm is especially useful for online prediction tasks. Another merit of the proposed BOLSSVM is that it can be used for k -fold cross validation. Specifically, compared with batch-mode learning methods, the computational complexity of the proposed BOLSSVM method is significantly reduced from O(n4) to O(n3) for leave-one-out cross validation with n training samples. The experimental results of classification and regression on benchmark data sets and real-world applications show the validity and effectiveness of the proposed BOLSSVM algorithm.

Adaptive natural gradient learning algorithms for Mackey–Glass chaotic time prediction

In this paper, we introduce the adaptive natural gradient method into the multilayer perceptrons (MLPs) and radial basis function (RBF) networks. We give a good performance for the Mackey–Glass chaotic time series prediction, and compare it with the LMA. Results show that the adaptive natural gradient methods for MLPs and RBFs, which are the online learning, can give almost the same performance with the LMA (MLPs), which is the batch mode learning. However, the performance of LMA (RBFs) is very poor and is very sensitive with the initial parameters.

Theoretical and numerical analysis of learning dynamics near singularity in multilayer perceptrons

The multilayer perceptron is one of the most widely used neural networks in applications, however, its learning behavior often becomes very slow, which is due to the singularities in the parameter space. In this paper, we analyze the learning dynamics near singularities in multilayer perceptrons by using traditional methods. We obtain the explicit expressions of the averaged learning equations which play a significant role in theoretical and numerical analysis. After obtaining the best approximation on overlap singularity, the stability of overlap singularity is analyzed. Then we take the numerical analysis on singular regions. Real averaged dynamics near the singularities are obtained in comparison with the theoretical learning trajectories near singularity. In the simulation we analyze the averaged learning dynamics, batch mode learning dynamics and on-line learning dynamics, respectively.

Online Social Spammer Detection

The explosive use of social media also makes it a popular platform for malicious users, known as social spammers, to overwhelm normal users with unwanted content. One effective way for social spammer detection is to build a classifier based on content and social network information. However, social spammers are sophisticated and adaptable to game the system with fast evolving content and network patterns. First, social spammers continually change their spamming content patterns to avoid being detected. Second, reflexive reciprocity makes it easier for social spammers to establish social influence and pretend to be normal users by quickly accumulating a large number of "human" friends. It is challenging for existing anti-spamming systems based on batch-mode learning to quickly respond to newly emerging patterns for effective social spammer detection. In this paper, we present a general optimization framework to collectively use content and network information for social spammer detection, and provide the solution for efficient online processing. Experimental results on Twitter datasets confirm the effectiveness and efficiency of the proposed framework.

Online Training for Open Faulty RBF Networks

Recently, a batch mode learning algorithm, namely optimal open weight fault regularization (OOWFR), was developed for handling the open fault situation. In terms of the Kullback---Leibler divergenc...

Sentiment Analysis on the Social Networks Using Stream Algorithms

The rising popularity of online social networks (OSNs), such as Twitter, Facebook, MySpace, and LinkedIn, in recent years has sparked great interest in sentiment analysis on their data. While many methods exist for identifying sentiment in OSNs such as communication pattern mining and classification based on emoticon and parts of speech, the majority of them utilize a suboptimal batch mode learning approach when analyzing a large amount of real time data. As an alternative we present a stream algorithm using Modified Balanced Winnow for sentiment analysis on OSNs. Tested on three real-world network datasets, the performance of our sentiment predictions is close to that of batch learning with the ability to detect important features dynamically for sentiment analysis in data streams. These top features reveal key words important to the analysis of sentiment.

Audio Segment Classification Using Online Learning Based Tensor Representation Feature Discrimination

In order to naturally combine audio information from different dimensions and build robust audio processing system, a novel framework based on low-rank tensor representation features for audio segment classification is proposed in this paper. The audio signal is first transformed into tensor format data, and then these tensor data are mapped to a low-rank space which is insensitive under certain noises, especially white Gaussian noise and gross corruptions. For these low-rank tensor based features, tensor classification via a linear classifier based on minimization a smooth loss function regularized by the trace norm proposed recently is used. Most previous methods find the weight tensor and bias in batch-mode learning, which makes them inefficient for large-scale problems. In this paper, we propose to address this problem with an online learning algorithm based on the accelerated proximal gradient (APG) method, which scales up gracefully to large data sets. Experiments on simulation and real audio data demonstrate the efficiency of the methods.

Fuzzy Associative Conjuncted Maps Network

The fuzzy associative conjuncted maps (FASCOM) is a fuzzy neural network that associates data of nonlinearly related inputs and outputs. In the network, each input or output dimension is represented by a feature map that is partitioned into fuzzy or crisp sets. These fuzzy sets are then conjuncted to form antecedents and consequences, which are subsequently associated to form if-then rules. The associative memory is encoded through an offline batch mode learning process consisting of three consecutive phases. The initial unsupervised membership function initialization phase takes inspiration from the organization of sensory maps in our brains by allocating membership functions based on uniform information density. Next, supervised Hebbian learning encodes synaptic weights between input and output nodes. Finally, a supervised error reduction phase fine-tunes the network, which allows for the discovery of the varying levels of influence of each input dimension across an output feature space in the encoded memory. In the series of experiments, we show that each phase in the learning process contributes significantly to the final accuracy of prediction. Further experiments using both toy problems and real-world data demonstrate significant superiority in terms of accuracy of nonlinear estimation when benchmarked against other prominent architectures and exhibit the network's suitability to perform analysis and prediction on real-world applications, such as traffic density prediction as shown in this paper.

Nuclear reactor dynamics on-line estimation by Locally Recurrent Neural Networks

In this paper, the Causal Recursive Back-Propagation (CRBP) algorithm is employed to train on-line an Infinite Impulse Response–Locally Recurrent Neural Network (IIR–LRNN) for modelling the dynamics of a next-generation nuclear reactor. The results demonstrate the advantages of the on-line training over the batch-mode learning in the reconstruction of complex nonlinear dynamic relationships.

Dynamics of learning near singularities in radial basis function networks

The radial basis function (RBF) networks are one of the most widely used models for function approximation in the regression problem. In the learning paradigm, the best approximation is recursively or iteratively searched for based on observed data (teacher signals). One encounters difficulties in such a process when two component basis functions become identical, or when the magnitude of one component becomes null. In this case, the number of the components reduces by one, and then the reduced component recovers as the learning process proceeds further, provided such a component is necessary for the best approximation. Strange behaviors, especially the plateau phenomena, have been observed in dynamics of learning when such reduction occurs. There exist singularities in the space of parameters, and the above reduction takes place at the singular regions. This paper focuses on a detailed analysis of the dynamical behaviors of learning near the overlap and elimination singularities in RBF networks, based on the averaged learning equation that is applicable to both on-line and batch mode learning. We analyze the stability on the overlap singularity by solving the eigenvalues of the Hessian explicitly. Based on the stability analysis, we plot the analytical dynamic vector fields near the singularity, which are then compared to those real trajectories obtained by a numeric method. We also confirm the existence of the plateaus in both batch and on-line learning by simulation.

Dichotomous Learning Algorithm for the Optimal Design of Weighted Order Statistics Filters

Weighted order statistics (WOS) filters are highly effective, in processing digital signals, due to their simple window structure. This paper proposes a fast and efficient learning algorithm that both improves learning speed and reduces the complexity of designing WOS filters. The algorithm uses a dichotomous approach to reduce the Boolean functions from 255 levels to two levels which are separated by an optimal hyperplane. The design concept of this algorithm is similar to that of support vector machines (SVMs), which use two separate sets of data to determine the optimal hyperplane. A conjugate gradient algorithm is adopted, to solve the orthant-constrained optimum problem, in order to improve memory storage for the large amounts of data required in the design process. Prior literature includes three different schemes for learning: one approach updates the parameters via pattern-mode learning, while the others involve batch-mode learning and semi-batch mode learning. Our proposed method approximates the optimal weighted order statistics filters far more rapidly than either Yoo's algorithm or adaptive neural filters.

Oblique Support Vector Machines

In this paper we propose a modified framework of support vector machines, called Oblique Support Vector Machines(OSVMs), to improve the capability of classification. The principle of OSVMs is joining an orthogonal vector into weight vector in order to rotate the support hyperplanes. By this way, not only the regularized risk function is revised, but the constrained functions are also modified. Under this modification, the separating hyperplane and the margin of separation are constructed more precise. Moreover, in order to apply to large-scale data problem, an iterative learning algorithm is proposed. In this iterative learning algorithm, three different schemes for training can be found in this literature, including pattern-mode learning, semi-batch mode learning and batch mode learning. Besides, smooth technique is adopted in order to convert the constrained nonlinear programming problem into unconstrained optimum problem. Consequently, experimental results and comparisons are given to demonstrate that the performance of OSVMs is better than that of SVMs and SSVMs.

An incremental EM-based learning approach for on-line prediction of hospital resource utilization

Inpatient length of stay (LOS) is an important measure of hospital activity, health care resource consumption, and patient acuity. This research work aims at developing an incremental expectation maximization (EM) based learning approach on mixture of experts (ME) system for on-line prediction of LOS. The use of a batch-mode learning process in most existing artificial neural networks to predict LOS is unrealistic, as the data become available over time and their pattern change dynamically. In contrast, an on-line process is capable of providing an output whenever a new datum becomes available. This on-the-spot information is therefore more useful and practical for making decisions, especially when one deals with a tremendous amount of data. The proposed approach is illustrated using a real example of gastroenteritis LOS data. The data set was extracted from a retrospective cohort study on all infants born in 1995-1997 and their subsequent admissions for gastroenteritis. The total number of admissions in this data set was n = 692. Linked hospitalization records of the cohort were retrieved retrospectively to derive the outcome measure, patient demographics, and associated co-morbidities information. A comparative study of the incremental learning and the batch-mode learning algorithms is considered. The performances of the learning algorithms are compared based on the mean absolute difference (MAD) between the predictions and the actual LOS, and the proportion of predictions with MAD < or = 1 day (Prop(MAD < or = 1)). The significance of the comparison is assessed through a regression analysis. The incremental learning algorithm provides better on-line prediction of LOS when the system has gained sufficient training from more examples (MAD = 1.77 days and Prop(MAD < or = 1) = 54.3%), compared to that using the batch-mode learning. The regression analysis indicates a significant decrease of MAD (p-value = 0.063) and a significant (p-value = 0.044) increase of Prop(MAD < or = 1) with the incremental learning algorithm. The incremental learning feature and the self-adaptive model-selection ability of the ME network enhance its effective adaptation to non-stationary LOS data. It is demonstrated that the incremental learning algorithm outperforms the batch-mode algorithm in the on-line prediction of LOS.

On 'natural' learning and pruning in multi-layered perceptrons.

Several studies have shown that natural gradient descent for on-line learning is much more efficient than standard gradient descent. In this article, we derive natural gradients in a slightly different manner and discuss implications for batch-mode learning and pruning, linking them to existing algorithms such as Levenberg-Marquardt optimization and optimal brain surgeon. The Fisher matrix plays an important role in all these algorithms. The second half of the article discusses a layered approximation of the Fisher matrix specific to multilayered perceptrons. Using this approximation rather than the exact Fisher matrix, we arrive at much faster "natural" learning algorithms and more robust pruning procedures.

Structurally adaptive modular networks for nonstationary environments

This paper introduces a neural network capable of dynamically adapting its architecture to realize time variant nonlinear input-output maps. This network has its roots in the mixture of experts framework but uses a localized model for the gating network. Modules or experts are grown or pruned depending on the complexity of the modeling problem. The structural adaptation procedure addresses the model selection problem and typically leads to much better parameter estimation. Batch mode learning equations are extended to obtain on-line update rules enabling the network to model time varying environments. Simulation results are presented throughout the paper to support the proposed techniques.

A theoretical comparison of batch-mode, on-line, cyclic, and almost-cyclic learning

We study and compare different neural network learning strategies: batch-mode learning, online learning, cyclic learning, and almost-cyclic learning. Incremental learning strategies require less storage capacity than batch-mode learning. However, due to the arbitrariness in the presentation order of the training patterns, incremental learning is a stochastic process; whereas batch-mode learning is deterministic. In zeroth order, i.e., as the learning parameter eta tends to zero, all learning strategies approximate the same ordinary differential equation for convenience referred to as the "ideal behavior". Using stochastic methods valid for small learning parameters eta, we derive differential equations describing the evolution of the lowest-order deviations from this ideal behavior. We compute how the asymptotic misadjustment, measuring the average asymptotic distance from a stable fixed point of the ideal behavior, scales as a function of the learning parameter and the number of training patterns. Knowing the asymptotic misadjustment, we calculate the typical number of learning steps necessary to generate a weight within order epsilon of this fixed point, both with fixed and time-dependent learning parameters. We conclude that almost-cyclic learning (learning with random cycles) is a better alternative for batch-mode learning than cyclic learning (learning with a fixed cycle).