Search

The discovery of process models from event logs has emerged as one of the crucial problems for enabling the continuous support in the life-cycle of an information system. However, in a decade of process discovery research, the algorithms and tools that have appeared are known to have strong limitations in several dimensions. The size of the logs and the formal properties of the model discovered are the two main challenges nowadays. In this paper we propose the use of numerical abstract domains for tackling these two problems, for the particular case of the discovery of Petri nets. First, numerical abstract domains enable the discovery of general process models, requiring no knowledge (e.g., the bound of the Petri net to derive) for the discovery algorithm. Second, by using divide and conquer techniques we are able to control the size of the process discovery problems. The methods proposed in this paper have been implemented in a prototype tool and experiments are reported illustrating the significance of this fresh view of the process discovery problem. [ABSTRACT FROM AUTHOR]

This paper presents a sensitivity analysis of similar days (SD) parameters to increase the accuracy of artificial neural network (ANN) and SD-based short-term price forecasting. Work presented in this paper is an extended version of previous works done by the authors to integrate ANN and SD method for predicting electricity price. The focus here is on sensitivity analysis of SD parameters while keeping the parameters same for ANN to forecast hourly electricity prices in the Pennsylvania--New Jersey--Maryland (PJM) (regional transmission organization in northeast America) electricity market. Several cases are simulated by choosing (a) two, (b) three, (c) four, and (d) five SD parameters to calculate the norm. In addition, sensitivity analysis has been carried out by changing the time framework of SD (d = 15,30, 45, 60) and the number of selected similar price days (N = 5, 10). From sensitivity analysis, it is identified that the optimized mean absolute percentage error (MAPE) is obtained using case-c with d = 30 and N = 10. MAPE of reasonably small value, along with forecast mean square error and mean absolute error of around 2$/MWh and 1$/MWh, is obtained for the PJM data, which has correlation coefficient of determination (R²) of 0.7758 between load and electricity price. Numerical results show that forecasts generated by the developed ANN model based on the optimized case are accurate and efficient. [ABSTRACT FROM AUTHOR]

This paper presents the module–phase synchronization for a class of neural networks with weights identification and time delays. In module–phase synchronization, complex-valued node states are taken into consideration. The topology weights considered here are uncertain and the time delays are bounded. By constructing a Lyapunov–Krasovskii functional and employing adaptive feedback control, sufficient conditions for module–phase synchronization are derived. After the general synchronization theory, the synchronization with topology identification is discussed. Finally, pertinent examples are given to demonstrate the effectiveness of the obtained results. [ABSTRACT FROM PUBLISHER]

This paper proposes a virtual 3D eyeglasses try-on system driven by a 2D Internet image of a human face wearing with a pair of eyeglasses. The main technical challenge of this system is the automatic 3D eyeglasses model reconstruction from the 2D glasses on a frontal human face. Against this challenge, this paper first proposes an eyeglasses segmentation method using a convolutional neural network-based parsing algorithm to label the glasses pixels, followed with a proposed symmetry-based level-set optimization algorithm to refine the contour of the eyeglasses. With the precisely extracted silhouette image, we take advantages of the smoothness and the symmetry priors of the eyeglasses, and propose a silhouette-based 3D deformation method to deform a 3D eyeglasses model selected from a predefined 3D model database. The obtained model is plausibly approximated to the input 2D eyeglasses after a texture mapping step. Finally, we develop a virtual try-on system to interactively synthesize the reconstructed eyeglasses on a moving target face in real time. The experimental results demonstrate the efficient and convincing virtual try-on performance of our approach and the commercial potential of our proposed system. [ABSTRACT FROM PUBLISHER]

The finite-time state estimation issue is addressed in this paper for discrete time-delayed neural networks (NNs). More than one communication channel is utilized to improve the communication performance. The transmission delays of each channel are modeled by a family of stochastic variables which are independent and identically distributed. The main purpose of this paper is to construct an appropriate state estimation scheme under which the corresponding state estimation error dynamics is finite-time bounded in the mean square. By employing the stochastic analysis approach and introducing a special Lyapunov-like functional, we have developed certain sufficient conditions to achieve the prescribed estimation performance. Furthermore, the exact expressions of the achieved estimator parameters are given by solving a special minimization problem subject to certain inequality constraints. Finally, we propose an illustrative simulation to examine the correctness, as well as the effectiveness, of our proposed state estimation method. [ABSTRACT FROM AUTHOR]

For a class of stochastic nonlinear systems in pure-feedback form with dead-zone input and multiple time-varying delays, a novel neural network (NN)-based adaptive control approach is presented in this paper through the use of backstepping approach and dynamic surface technique. By choosing proper Lyapunov–Krasovskii functionals, utilizing the characteristic of hyperbolic tangent functions and adopting the function separation technique, difficulties of controller design that introduced by the time-varying delays can be dealt with properly. Moreover, all unknown nonlinear functions are lumped together and approximated by the NN. Additionally, any information over the boundedness of dead-zone parameters is not needed in the process of controller design. The control scheme proposed in this paper ensures the boundedness in probability of all signals in the closed-loop system, besides, excellent performance of arbitrarily small tracking error will be achieved by selecting control parameters appropriately. At last, two numerical simulation examples are provided to verify the validity of the designed algorithm. [ABSTRACT FROM AUTHOR]

This paper establishes an approximate optimal critic learning algorithm based on single neural network (NN) policy iteration (PI) aiming at solving for continuous-time (CT) 2-player zero-sum games (ZSGs). In fact, we have to face the problem that the errors will disturb the dynamics and in turn identifying dynamics will generate errors. In order to prevent the effect of errors, in this paper, a single NN-based online PI algorithm is developed for the CT system, which is disturbed nonlinear ZSG. With plenty of online data, the Hamilton–Jacobi–Isaacs equation can be solved without complete dynamics. Then by the least-squares method, we can obtain the NN weights. Moreover, in the process of dealing with the undisturbed system, we find the way that obtains NN weights in this paper is equal to the way that obtains the optimal solution by the Gauss–Newton method. Based on the convergence of the Gauss–Newton method, we can efficiently obtain the optimal controller for the undisturbed system by utilizing online data. After getting the controller of the undisturbed system, it is time to take disturbance into consideration, so that we design a robust control pair to overcome the disturbance. In order to demonstrate the effectiveness of this algorithm, we design a set of simulations. The results verify that we can solve the disturbed nonlinear ZSG by this algorithm. [ABSTRACT FROM AUTHOR]

Deep neural network (DNN) models have been widely used in many tasks due to their superior performance. However, DNN models are usually vulnerable to adversarial example attacks, which limits their applications in many safety-critic scenarios. How to effectively detect adversarial examples to enhance the robustness of DNN models has attracted much attention in recent years. Most adversarial example detection methods require modifying or retraining the model, which is impractical and reduces the classification accuracy of normal examples. In this paper, we propose an adversarial example detection approach that does not require modification of the DNN models and meanwhile retains the classification accuracy of normal examples. The key observation is that when we transform the input example with some operations (e.g., masking a pixel with a reference value), feed the transformed example to the target model, and use the output of the intermediate layers to predict the label of the example, the generated label sequences of adversarial examples will be extremely discrepant but the label sequences of normal examples keep nearly unchanged. Motivated by this observation, we design an approach to detect adversarial examples based on the label sequence discrepancy (LSD) of the given examples. The experimental results against five mainstream adversarial attacks on three benchmark datasets demonstrate that LSD outperforms the state-of-the-art solutions in the detection rate of adversarial examples. Moreover, LSD performs well at various confidence levels and exhibits good generalizability between different attacks. [ABSTRACT FROM AUTHOR]

Wi-Fi fingerprinting is a well-known technique used for indoor positioning. It relies on a pattern recognition method that compares the captured operational fingerprint with a set of previously collected reference samples (radio map) using a similarity function. The matching algorithms suffer from a scalability problem in large deployments with a huge density of fingerprints, where the number of reference samples in the radio map is prohibitively large. This paper presents a comprehensive comparative study of existing methods to reduce the complexity and size of the radio map used at the operational stage. Our empirical results show that most of the methods reduce the computational burden at the expense of a degraded accuracy. Among the studied methods, only $k$ k -means, affinity propagation, and the rules based on the strongest access point properly balance the positioning accuracy and computational time. In addition to the comparative results, this paper also introduces a new evaluation framework with multiple datasets, aiming at getting more general results and contributing to a better reproducibility of new proposed solutions in the future. [ABSTRACT FROM AUTHOR]

Neural networks (NNs) are attractive for nonlinear impairment compensation applications in communication systems, such as optical fiber nonlinearity, nonlinearity of driving amplifiers, and nonlinearity of semiconductor optical amplifiers. Without prior knowledge of the transmission link or the hardware characteristics, optimal parameters are completely constructed from a data-driven approach by exploring training datasets, once the NN structure is given. On the other hand, due to computational power and energy consumption, especially in high-speed communication systems, the computational complexity of the optimized NN needs to be confined to the hardware, such as FPGA or ASIC without sacrificing its performance improvement. In this paper, two approaches are presented to accommodate the NN-based algorithms for high-speed communication systems. The first approach is to reduce computational complexity of the NN-based nonlinearity compensation algorithms on the basis of weight pruning (WP). WP can significantly reduce the computational complexity, especially because the nonlinear compensation task studied here results in a sparse NN. The authors have studied an enhanced approach of WP by imposing an additional restriction on the selection of non-zero weights on each hidden layer. The second approach is to implement NNs onto a silicon-photonic integrated platform, enabling power efficiency to be further improved without sacrificing the high-speed operation. [ABSTRACT FROM AUTHOR]

Using deep neural networks (DNNs) to explore spatial patterns and temporal dynamics of human brain activities has been an important yet challenging problem because the artificial neural networks are hard to be designed manually. There have been several promising deep learning methods, e.g., deep belief network (DBN), convolutional neural network (CNN), and deep sparse recurrent auto-encoder (DSRAE), that can decompose neuroscientific and meaningful spatiotemporal patterns from 4D functional Magnetic Resonance Imaging (fMRI) data. However, those previous studies still depend on hand-crafted neural network architectures and hyperparameters, which are not optimal in various senses. In this paper, we employ the evolutionary algorithms (EA) to optimize the deep neural architecture of DSRAE by minimizing the expected loss of initialized models, named eNAS-DSRAE (evolutionary Neural Architecture Search on Deep Sparse Recurrent Auto-Encoder). Also, validation experiments are designed and performed on the publicly available human connectome project (HCP) 900 datasets, and the results achieved by the optimized eNAS-DSRAE suggested that our framework can successfully identify the spatiotemporal features and perform better than the hand-crafted neural network models. To our best knowledge, the proposed eNAS-DSRAE is not only among the earliest NAS models that can extract connectome-scale meaningful spatiotemporal brain networks from 4D fMRI data, but also is an effective framework to optimize the RNN-based models. [ABSTRACT FROM AUTHOR]

Inter prediction is a crucial part of hybrid video coding frameworks, utilized to exploit the temporal redundancy in video sequences and improve the coding performance. During inter prediction, a predicted block is typically derived from reference pictures using motion estimation and motion compensation. To improve the coding performance of inter prediction, a neural network based enhancement to inter prediction (NNIP) is proposed in this paper. NNIP is composed of three networks, namely residue estimation network, combination network, and deep refinement network. Specifically, first, a residue estimation network is designed to estimate the residue between current block and its predicted block using their available spatial neighbors. Second, the feature maps of the estimated residue and the predicted block are extracted and concatenated in a combination network. Finally, the concatenated feature maps are fed into a deep refinement network to generate a refined residue, which is added back to the predicted block to derive a more accurate predicted block. NNIP is integrated in HEVC to evaluate its efficiency. The experimental results demonstrate that NNIP can achieve 4.6%, 3.0%, and 2.7% BD-rate reduction on average under LDP, LDB, and RA configurations compared to HEVC. [ABSTRACT FROM AUTHOR]

High-symbol-rate coherentoptical transceivers suffer more from the critical responses of transceiver components at high frequency, especially when applying a higher order modulation format. We recently proposed a neural network (NN)-based digital pre-distortion (DPD) technique trained to mitigate the transceiver response of a 128 GBaud optical coherent transmission system. In this paper, we further detail this work and assess the NN-based DPD by training it using either a direct learning architecture (DLA) or an indirect learning architecture (ILA), and compare performance against a Volterra series-based ILA DPD and a linear DPD. Furthermore, we deliberately increase the transmitter nonlinearity and compare the performance of the three DPDs schemes. The proposed NN-based DPD trained using DLA performs the best among the three contenders. In comparison to a linear DPD, it provides more than 1 dB signal-to-noise ratio (SNR) gains at the output of a conventional coherent receiver DSP for uniform 64-quadrature amplitude modulation (QAM) and PCS-256-QAM signals. Finally, the NN-based DPD enables achieving a record 1.61 Tb/s net rate transmission on a single channel after 80 km of standard single mode fiber (SSMF). [ABSTRACT FROM AUTHOR]

Massive multiple-input multiple-output can obtain more performance gain by exploiting the downlink channel state information (CSI) at the base station (BS). Therefore, studying CSI feedback with limited communication resources in frequency-division duplexing systems is of great importance. Recently, deep learning (DL)-based CSI feedback has shown considerable potential. However, the existing DL-based explicit feedback schemes are difficult to deploy because current fifth-generation mobile communication protocols and systems are designed based on an implicit feedback mechanism. In this paper, we propose a DL-based implicit feedback architecture to inherit the low-overhead characteristic, which uses neural networks (NNs) to replace the precoding matrix indicator (PMI) encoding and decoding modules. By using environment information, the NNs can achieve a more refined mapping between the precoding matrix and the PMI compared with codebooks. The correlation between subbands is also used to further improve the feedback performance. Simulation results show that, for a single resource block (RB), the proposed architecture can save 25.0% – 40.0% of overhead compared with the Type I codebook under different antenna configurations. For a wideband system with 52 RBs, overhead can be saved by 30.7% and 48.0% compared with the Type II codebook when ignoring and considering extracting subband correlation, respectively. [ABSTRACT FROM AUTHOR]

Recently, evolving fuzzy systems have been proved to be effective in dealing with real-time data streams. However, their fixed structures are not flexible enough to address the structural variations triggered by the changing operating conditions or system states in complex industrial environments. A novel generalized heterogeneous interval type-2 (IT2) fuzzy classifier, named as GHIT2Class, is proposed in this paper, which is built upon a multivariable IT2 fuzzy neural network. To fully reflect the industrial data characteristics of uncertainty, this paper proposes an approach of constructing the uncertainty footprint with ellipsoidal rotation. A rule pruning method based on error and incentive intensity dynamic adjustment mechanism is reported in the process of modeling, and a corresponding rule recall mechanism is designed to avoid rules of catastrophic forgetting. In addition, the simultaneous update of the upper and lower bounds of IT2 fuzzy consequent parameters is designed to relieve the computing overhead of the fuzzy systems. The performance of the proposed GHIT2Class is experimentally validated by a number of synthetic datasets and industry study cases by using state-of-the-art comparative classifiers, where the proposed approach outperforms the others in achieving the best tradeoff between accuracy and simplicity. [ABSTRACT FROM AUTHOR]

This paper investigates the synchronization problem of Takagi–Sugeno fuzzy memristive neural networks (FMNNs) with mixed delays, in which the bounded distributed and unbounded discrete time-varying delays are involved. Then, under the nonsmooth analysis and Lyapunov stability theory, several easily verified algebraic criteria are established to guarantee the global synchronization of FMNNs via a designed fuzzy feedback controller. Moreover, to show the superiority of the theoretical results, several discussions and comparisons with existing work are provided, indicating that derived results in this paper are general and include several existing ones as special cases. Finally, two numerical examples and two applications in psuedorandom number generation and image encryption are presented to show the validity and practicability of the theoretical results. [ABSTRACT FROM AUTHOR]

In recent years, the tracking problem of spherical inverted pendulum (SIP) system with multi-input, multi-output nature and unstable zero dynamics has been well addressed based on continuous-time nonlinear output regulation (NOR) theory. For the convenience of digital implementation, this paper further investigates the approximate NOR problem of the SIP system in discrete-time framework. The key for solving the discrete-time NOR problem lies in how to solve a set of algebraic functional equations known as discrete regulator equations (DRE). Since the equations are very complicated, the accurate solution of the DRE can not be obtained. In this paper, we first show that the DRE associated with the SIP system are solvable by center manifold theorem and then use neural network approach to tackle with the tracking problem. Finally, we compare our method with polynomial approximation method. [ABSTRACT FROM AUTHOR]

Image retargeting techniques adjust images into different sizes and have attracted much attention recently. Objective quality assessment (OQA) of image retargeting results is often desired to automatically select the best results. Existing OQA methods train a model using some benchmarks (e.g., RetargetMe), in which subjective scores evaluated by users are provided. Observing that it is challenging even for human subjects to give consistent scores for retargeting results of different source images (diff-source-results), in this paper we propose a learning-based OQA method that trains a General Regression Neural Network (GRNN) model based on relative scores—which preserve the ranking—of retargeting results of the same source image (same-source-results). In particular, we develop a novel training scheme with provable convergence that learns a common base scalar for same-source-results. With this source specific offset, our computed scores not only preserve the ranking of subjective scores for same-source-results, but also provide a reference to compare the diff-source-results. We train and evaluate our GRNN model using human preference data collected in RetargetMe. We further introduce a subjective benchmark to evaluate the generalizability of different OQA methods. Experimental results demonstrate that our method outperforms ten representative OQA methods in ranking prediction and has better generalizability to different datasets. [ABSTRACT FROM AUTHOR]

Based on the knowledge accumulated off-line and feature extractions, a novel data-based learning and control method is proposed for the long-term voltage stability problem in this paper. All the spatial-temporal data is considered and the features of different emergency events are extracted by principle component analysis which can reduce the dimension and reveal the significant internal structure of the data. An artificial neural network is used to build a classifier to reinforce the relationship directly between the system dynamics and optimal control actions. With the prepared control knowledge, it is faster to find an optimal control action online with a good system performance. Simulation results on the 6-bus system, New England 39-bus system and Iceland 189-bus system are given to show the potential of this method for on-line control. [ABSTRACT FROM AUTHOR]

Efficient nonlinearity compensation in fiber-optic communication systems is considered a key element to go beyond the “capacity crunch”. One guiding principle for previous work on the design of practical nonlinearity compensation schemes is that fewer steps lead to better systems. In this paper, we challenge this assumption and show how to carefully design multi-step approaches that provide better performance–complexity trade-offs than their few-step counterparts. We consider the recently proposed learned digital backpropagation (LDBP) approach, where the linear steps in the split-step method are re-interpreted as general linear functions, similar to the weight matrices in a deep neural network. Our main contribution lies in an experimental demonstration of this approach for a 25 Gbaud single-channel optical transmission system. It is shown how LDBP can be integrated into a coherent receiver DSP chain and successfully trained in the presence of various hardware impairments. Our results show that LDBP with limited complexity can achieve better performance than standard DBP by using very short, but jointly optimized, finite-impulse response filters in each step. This paper also provides an overview of recently proposed extensions of LDBP and we comment on potentially interesting avenues for future work. [ABSTRACT FROM AUTHOR]

The classification of distinct problems of insulators in the distribution networks is a task that requires operator's experience. The applications of techniques to automate the inspection of electrical systems with the objective of detecting faults in insulators have shown to be reasonable alternatives to improve reliability in power grid. In this paper, based on the development of an experimental setup, signals are acquired considering three distinct faults in insulators. In this case, 13.8 kV (rms) is applied in drilled, contaminated, and good insulators considering an ultrasound detector connected to a computer. In the sequence, a multiclass classification method is proposed considering the ensemble of classifiers. The method considers the association of five distinct techniques, Bottom-Up segmentation, wavelet energy coefficient, principal component analysis, and particle swarm optimization associated with ensemble extreme learning machine (EN-ELM). Named optimized ensemble extreme learning machine, the present approach outperforms the original EN-ELM method. Finally, results show significant increase in robustness and faster training procedure when compared to classical approaches. [ABSTRACT FROM AUTHOR]

As the failure of power line insulators leads to the failure of power transmission systems, an insulator inspection system based on an aerial platform is widely used. Insulator defect detection is performed against complex backgrounds in aerial images, presenting an interesting but challenging problem. Traditional methods, based on handcrafted features or shallow-learning techniques, can only localize insulators and detect faults under specific detection conditions, such as when sufficient prior knowledge is available, with low background interference, at certain object scales, or under specific illumination conditions. This paper discusses the automatic detection of insulator defects using aerial images, accurately localizing insulator defects appearing in input images captured from real inspection environments. We propose a novel deep convolutional neural network (CNN) cascading architecture for performing localization and detecting defects in insulators. The cascading network uses a CNN based on a region proposal network to transform defect inspection into a two-level object detection problem. To address the scarcity of defect images in a real inspection environment, a data augmentation method is also proposed that includes four operations: 1) affine transformation; 2) insulator segmentation and background fusion; 3) Gaussian blur; and 4) brightness transformation. Defect detection precision and recall of the proposed method are 0.91 and 0.96 using a standard insulator dataset, and insulator defects under various conditions can be successfully detected. Experimental results demonstrate that this method meets the robustness and accuracy requirements for insulator defect detection. [ABSTRACT FROM AUTHOR]

Although high dynamic range (HDR) display has become popular recently, the legacy content such as standard dynamic range (SDR) video is still in service and needs to be properly converted on HDR display devices. Therefore, it is desirable for HDR TV sets to have the capability of automatically converting input SDR video into HDR video, which is called reverse tone mapping (RTM). In this paper, we propose a novel learning-based low-complexity RTM scheme that not only expands the suppressed dynamic ranges (DR) of the SDR videos (or images), but also effectively restores lost detail in the SDR videos. Most existing conventional RTM schemes have focused on how to expand the DR of global contrast, resulting in limitations in recovering lost detail of SDR videos. On the other hand, the recent convolutional neural network-based approaches show promising results, but they are too complex to be applied on the users’ devices in practice. In this paper, our learning-based RTM scheme is computationally simple but effective in recovering lost detail. To learn the SDR-to-HDR relation, training “SDR-HDR” images are first separated into their base layer components and detail layer components by applying a guided filter. The detail layer components of the “SDR-HDR” pairs are used to train the SDR-to-HDR mapping. The mapping matrices are computed based on kernel ridge regression. In the meantime, the global contrast of the base layers is expanded by a nonlinear function that suppresses darker regions and amplifies brighter regions to fit the full DR of a target HDR display. To verify the effectiveness of our learning-based RTM scheme, we performed subjective quality assessment for images and videos. The experimental results show that our RTM scheme outperforms the existing RTM scheme with the successful restoration of lost detail in SDR images. [ABSTRACT FROM AUTHOR]