Search

Power system inertia is declining and is increasingly variable and uncertain in regions where the penetration of non-synchronous generation and interconnectors is growing. This presents a challenge to power system operators who must take appropriate actions to ensure the stability and security of power systems relying on short-term forecasts of the system’s inertial response. Existing models to forecast inertia fail to quantify uncertainty, which may prevent their utilization given the risk aversion of the system operators when handling stability issues. This paper is the first to develop a model to produce calibrated, data-driven probabilistic forecasts of the inertia contribution of transmission-connected synchronous generators. The model provides a necessary tool for system operators to quantify forecast uncertainty, allowing them to manage the risk of frequency instability cost-effectively. The paper demonstrates that the assumption of a Gaussian distribution of uncertainty applied in existing models is not acceptable to accurately forecast the inertial response and provides a satisfactory forecast model by combining non-parametric density forecasting with parametric tail distributions. Moreover, the paper shows that satisfactory predictive performance can only be achieved by adopting a rolling horizon forecast approach to deal with the rapidly changing characteristics of the inertial response in power systems. [ABSTRACT FROM AUTHOR]

Load profile classification is very important in load forecast, planning and management. Although customers are generally grouped by utilities into residential, commercial classes and respective subclasses, there is a lack of systematic framework that can be used to characterize different classes with signatures that are both human-readable and machine-readable. The work presented in this paper attempts to formulate the theoretical framework for customer classification using the annual load profiles. This paper demonstrates how to extract characteristic attributes in frequency domain (CAFD) and use these CAFDs to formulate a hierarchy of load profiles that can be used as the systematic framework for customer load classification. As signatures for customer classes and subclasses, the CAFDs are obtained by using a data mining method called CART (classification and regression tree). The paper presents a load profile classification test to establish the efficacy of the proposed approach which is significant improvement over current practices that provide mostly qualitative labeling. [ABSTRACT FROM AUTHOR]

The remaining useful life (RUL) of transformer insulation paper is largely determined by the winding hot-spot temperature (HST). Frequently the HST is not directly monitored and it is inferred from other measurements. However, measurement errors affect prediction models and if uncertain variables are not taken into account this can lead to incorrect maintenance decisions. Additionally, existing analytic models for HST calculation are not always accurate because they cannot generalize the properties of transformers operating in different contexts. In this context, this paper presents a novel transformer condition assessment approach integrating uncertainty modeling, data-driven forecasting models, and model-based experimental models to increase the prediction accuracy and handle uncertainty. The proposed approach quantifies the effect of measurement errors on transformer RUL predictions and confirms that temperature and load measurement errors affect the RUL estimation. Forecasting results show that the extreme gradient boosting (XGB) algorithm best captures the nonlinearities of the thermal model and improves the prediction accuracy among a number of forecasting approaches. Accordingly, the XGB model is integrated with experimental models in a particle filtering framework to improve thermal modeling and RUL prediction tasks. Models are tested and validated using a real dataset from a power transformer operating in a nuclear power plant. [ABSTRACT FROM AUTHOR]

In the field of reversible data hiding (RDH), how to predict an image and embed a message into the image with smaller distortion are two important aspects. In this paper, we propose a novel and efficient RDH method by innovating an intelligent predictor and an adaptive embedding way. In the prediction stage, we first constructed a convolutional neural network (CNN) based predictor by reasonably dividing an image into four parts. In such a way, each part can be predicted by using the other three parts as the context for the improvement of the prediction performance. Compared with existing predictors, the proposed CNN predictor can use more neighboring pixels for the prediction by exploiting its multi-receptive fields and global optimization capacities. In the embedding stage, we also developed a prediction-error-ordering (PEO) based adaptive embedding strategy, which can better adapt image content and thus efficiently reduce the embedding distortion by elaborately and luminously applying background complexity to select and pair those smaller prediction errors for data hiding. With the proposed CNN prediction and embedding ways, the RDH method presented in this paper provides satisfactory results in improving the visual quality of data hidden images, e.g., the average PSNR value for the Kodak benchmark dataset can reach as high as 63.59 dB with an embedding capacity of 10,000 bits. Extensive experimental results have shown that the RDH method proposed in this paper is superior to those existing state-of-the-art works. [ABSTRACT FROM AUTHOR]

Network embedding has been widely employed in networked data mining applications as it can learn low-dimensional and dense node representations from the high-dimensional and sparse network structure. While most existing network embedding methods only model the proximity between two nodes regardless of the order of the proximity, this paper proposes to explicitly model multi-node proximities which can be widely observed in practice, e.g., multiple researchers coauthor a paper, and multiple genes co-express a protein. Explicitly modeling multi-node proximities is important because some two-node interactions may not come into existence without a third node. By proving that LINE(1st), a recent network embedding method, is equivalent to kernelized matrix factorization, this paper proposes coupled kernelized multi-dimensional array factorization (Cetera) which jointly factorizes multiple multi-dimensional arrays by enforcing a consensus representation for each node. In this way, node representations can be more comprehensive and effective, which is demonstrated on three real-world networks through link prediction and multi-label classification. [ABSTRACT FROM AUTHOR]

This paper presents an asset health index (HI) prediction methodology for high voltage transmission overhead lines (OHLs) using supervised machine learning and structured, unambiguous visual inspections. We propose a framework for asset HI predictions to determine the technical condition of individual OHL towers to improve grid reliability in a cost-effective manner. The paper focuses on asset HI prediction and the selection of the most parsimonious model. Based on the technical specifications and HI data, our methodology allows for the prediction of a HI for OHLs without HI data, and models asset aging behaviour. Technical specifications and the HI as defined in this paper are taken from the Estonian TSO periodical visual inspections implemented in 2018. The case study successfully demonstrates that the proposed methodology can predict tower HI values for a single OHL with nearly 80 percent accuracy without the need for additional measurements. [ABSTRACT FROM AUTHOR]

Accurate load forecasting is the key to economic dispatch and efficient operation of Multi-Energy System (MES). This paper proposes a combined load forecasting method for MES based on Bi-directional Long Short-Term Memory (BiLSTM) multi-task learning. Firstly, this paper investigates the multi-energy interaction mechanism and multi-loads characteristics and analyzes the correlation of multi-loads in different seasons. Then, a combined load forecasting method is proposed, which focuses on making full use of the coupling relationship among multiple loads. In the forecasting model, the different loads are selected combinedly as the input features according to the Maximum Information Coefficient (MIC). The multi-task learning is adopted to construct the cooling, heating and electric combined load forecasting model based on the BiLSTM algorithm, which can effectively share the coupling information among the loads. Finally, case studies verify the effectiveness and superiority of the proposed method in both learning speed and forecasting accuracy. [ABSTRACT FROM AUTHOR]

With the development of Internet technology, information has become an increasingly important asset for the operation and planning of power systems. However, the existing studies and practices pay little attention to information value evaluation, i.e., the potential for information to be converted into actual economic benefits. To this end, this paper designs an information market framework and proposes a generalized information valuation model to help price data in smart grids efficiently. Here we analyze the information value of photovoltaic (PV)-related data in a power system operation problem. Specifically, we examine how additional meteorological and PV power data help to improve day-ahead forecasting accuracy, thus enhancing unit commitment (UC). In this paper, information quality is captured by two indices of a set of PV-related data, i.e., Shannon entropy and non-noise ratio. Then a neural network-based engine is employed to predict day-ahead hourly solar power on the premise of input datasets with different information quality. Here we define forecasting accuracy as information utility, and discover an exponential relationship between such utility and information quality. Finally, a two-stage stochastic UC model is formulated to quantify the contributions of different PV-related datasets, in which real-time solar power deviation is penalized. In this instance, the economic value of PV-related data is measured as the operational cost reduction induced by forecasting accuracy improvement, which we find can be estimated by information quality. Case studies based on the IEEE 30- and 118-bus systems validate the effectiveness of the proposed paradigm and method. [ABSTRACT FROM AUTHOR]

The dielectric modulus Μ∗(ω) is used for studying the frequency domain response and activation energy (E a) of transformer immersed cellulose insulation. Existing literatures show that the temperature is a crucial factor affecting the Μ∗(ω) and aging rate of transformer oil-paper insulation, while the E a can be regarded as an indicator to reflect the average aging rate of the transformer immersed cellulose insulation. Therefore, the discussion of the temperature-dependence on both Μ∗(ω) and E a is of great significance. This paper attempts to report an approach for understanding the temperature impact on both Μ∗(ω) and E a. The obtained knowledge can be used to establish two available models for temperature correction on Μ∗(ω) and activation energy prediction of transform immersed cellulose insulation, respectively. These findings are expected to promote the reliability of the condition prediction result of transformer immersed cellulose insulation by using Μ∗(ω) and E a technique. [ABSTRACT FROM AUTHOR]

Tidal power is an emerging technology with great potential to provide a sustainable means of renewable energy in many areas worldwide. However, the nature of the underwater environment provides challenges. Submerged machinery cannot be easily accessed for inspections, and turbines must be brought to the surface for maintenance. This is an expensive process and results in prolonged periods of downtime where no power can be supplied to the grid. Condition monitoring systems, capable of accurately and remotely assessing the health state of machinery while in operation, can therefore be of great value to this industry. This paper presents an approach for condition monitoring of a tidal turbine’s gearbox from monitoring data with low sample rates. Models of normal behavior were trained using weighted least squares regression, where prediction errors are used to identify changes in response. This paper then examines how prediction errors from a number of different cases (including changes in control scheme and simulated gearbox faults) can be interpreted by operators to classify anomalous behavior. [ABSTRACT FROM AUTHOR]

Link prediction aims at inferring missing links or predicting future ones based on the currently observed network. This topic is important for many applications such as social media, bioinformatics and recommendation systems. Most existing methods focus on homogeneous settings and consider only low-order pairwise relations while ignoring either the heterogeneity or high-order complex relations among different types of nodes, which tends to lead to a sub-optimal embedding result. This paper presents a method named Heterogeneous Hypergraph Variational Autoencoder (HeteHG-VAE) for link prediction in heterogeneous information networks (HINs). It first maps a conventional HIN to a heterogeneous hypergraph with a certain kind of semantics to capture both the high-order semantics and complex relations among nodes, while preserving the low-order pairwise topology information of the original HIN. Then, deep latent representations of nodes and hyperedges are learned by a Bayesian deep generative framework from the heterogeneous hypergraph in an unsupervised manner. Moreover, a hyperedge attention module is designed to learn the importance of different types of nodes in each hyperedge. The major merit of HeteHG-VAE lies in its ability of modeling multi-level relations in heterogeneous settings. Extensive experiments on real-world datasets demonstrate the effectiveness and efficiency of the proposed method. [ABSTRACT FROM AUTHOR]

In this paper, we study the problem of estimating subjective visual properties (SVP) for images, which is an emerging task in Computer Vision. Generally speaking, collecting SVP datasets involves a crowdsourcing process where annotations are obtained from a wide range of online users. Since the process is done without quality control, SVP datasets are known to suffer from noise. This leads to the issue that not all samples are trustworthy. Facing this problem, we need to develop robust models for learning SVP from noisy crowdsourced annotations. In this paper, we construct two general robust learning frameworks for this application. Specifically, in the first framework, we propose a probabilistic framework to explicitly model the sparse unreliable patterns that exist in the dataset. It is noteworthy that we then provide an alternative framework that could reformulate the sparse unreliable patterns as a “contraction” operation over the original loss function. The latter framework leverages not only efficient end-to-end training but also rigorous theoretical analyses. To apply these frameworks, we further provide two models as implementations of the frameworks, where the sparse noise parameters could be interpreted with the HodgeRank theory. Finally, extensive theoretical and empirical studies show the effectiveness of our proposed framework. [ABSTRACT FROM AUTHOR]

The intelligent radio is a novel emerging paradigm catching on in many practical applications, in order to provide context awareness techniques, able to learn by the environment conditions applying machine learning frameworks. The channel state information (CSI) prediction is one of the main challenges that an intelligent radio has to face with. This paper deals with the CSI prediction in the case of a buffer aided relaying communication scheme with adaptive link selection previously proposed in the literature where a direct source-destination link is not available and with both links affected by an independent fading. In particular, the CSI prediction has been reduced here to a time series prediction problem, performed by involving the usage of the recurrent neural networks. Finally, the validation of the effectiveness of the proposed intelligent radio buffer-aided relaying scheme with adaptive link selection approach has been pursued by presenting performance comparisons with the previously proposed scheme under the assumption of an ideal and actual CSI estimations, respectively. [ABSTRACT FROM AUTHOR]

This paper deals with randomized polling of a social network. In the case of forecasting the outcome of an election between two candidates A and B, classical intent polling asks randomly sampled individuals: who will you vote for? Expectation polling asks: who do you think will win? In this paper, we propose a novel neighborhood expectation polling (NEP) strategy that asks randomly sampled individuals: what is your estimate of the fraction of votes for A? Therefore, in NEP, sampled individuals will naturally look at their neighbors (defined by the underlying social network graph) when answering this question. Hence, the mean squared error (MSE) of NEP methods rely on selecting the optimal set of samples from the network. To this end, we propose two NEP algorithms for the following cases: (i) the social network graph is not known but, random walks (sequential exploration) can be performed on the graph, and (ii) the social network graph is unknown but, uniformly sampled nodes from the network are available. For both cases, algorithms based on a graph theoretic consequence called friendship paradox are proposed. Theoretical results on the dependence of the MSE of the algorithms on the properties of the network are established. Numerical results on real and synthetic data sets are provided to illustrate the performance of the algorithms. [ABSTRACT FROM AUTHOR]

Over the past several years, we have witnessed impressive progress in the field of learned image compression. Recent learned image codecs are commonly based on autoencoders, that first encode an image into low-dimensional latent representations and then decode them for reconstruction purposes. To capture spatial dependencies in the latent space, prior works exploit hyperprior and spatial context model to build an entropy model, which estimates the bit-rate for end-to-end rate-distortion optimization. However, such an entropy model is suboptimal from two aspects: (1) It fails to capture global-scope spatial correlations among the latents. (2) Cross-channel relationships of the latents remain unexplored. In this paper, we propose the concept of separate entropy coding to leverage a serial decoding process for causal contextual entropy prediction in the latent space. A causal context model is proposed that separates the latents across channels and makes use of channel-wise relationships to generate highly informative adjacent contexts. Furthermore, we propose a causal global prediction model to find global reference points for accurate predictions of undecoded points. Both these two models facilitate entropy estimation without the transmission of overhead. In addition, we further adopt a new group-separated attention module to build more powerful transform networks. Experimental results demonstrate that our full image compression model outperforms standard VVC/H.266 codec on Kodak dataset in terms of both PSNR and MS-SSIM, yielding the state-of-the-art rate-distortion performance. [ABSTRACT FROM AUTHOR]

With the aim of synchronizing high-speed railway (HSR) and aviation services to adapt intermodality traffic needs, this paper is concerned with HSR-Air timetable coordination (HATC) problem. This problem is solved by rescheduling the HSR timetable with the goal of attracting the maximum HSR-Air passenger flow, and at the same time, considering the minimum adjustments to the initial HSR timetable. The HATC problem is an integration of train timetable rescheduling and HSR-air passenger flow predicting problem. There is a trade-off increasing the HSR-Air passenger flow and the adjustments to the initial train timetable. In order to capture the complex feature interactions of passenger flow impact features, a novel HSR-Air passenger flow prediction model is proposed by using factorization machine and deep neural networks in this paper. Moreover, this passenger flow prediction model then is integrated into the train timetable rescheduling model to calculate the passenger flow under different HSR-Air service network. An approach based on a genetic algorithm is developed to solve the integrated model of deep learning and integer programming. The model and approach are tested in a real-world HSR-Air case in China with 15 HSR stations, 200 trains and 82 flights. The results show that the proposed model can obtain the satisfactory predictions and effectively enhance the HSR-Air passenger flow within an acceptable level of deviations to initial train timetable. [ABSTRACT FROM AUTHOR]

This paper focuses on channel prediction techniques for massive multiple-input multiple-output (MIMO) systems. Previous channel predictors are based on theoretical channel models, which would be deviated from realistic channels. In this paper, we develop and compare a vector Kalman filter (VKF)-based channel predictor and a machine learning (ML)-based channel predictor using the realistic channels from the spatial channel model (SCM), which has been adopted in the 3GPP standard for years. First, we propose a low-complexity mobility estimator based on the spatial average using a large number of antennas in massive MIMO. The mobility estimate can be used to determine the complexity order of developed predictors. The VKF-based channel predictor developed in this paper exploits the autoregressive (AR) parameters estimated from the SCM channels based on the Yule-Walker equations. Then, the ML-based channel predictor using the linear minimum mean square error (LMMSE)-based noise pre-processed data is developed. Numerical results reveal that both channel predictors have substantial gain over the outdated channel in terms of the channel prediction accuracy and data rate. The ML-based predictor has larger overall computational complexity than the VKF-based predictor, but once trained, the operational complexity of ML-based predictor becomes smaller than that of VKF-based predictor. [ABSTRACT FROM AUTHOR]

Currently, statistical time-series methods have primarily been employed to predict time-aware dynamic quality of service (QoS) attributes for Web services. In this paper, we propose the application of genetic programming (GP) for such predictions. Our experimental results indicate that the GP-based approach is more accurate than the other approaches presented for comparison. However, for the efficient management of such attributes for cloud services, including their modeling and forecasting, the current research is insufficient because a set of research questions remains unanswered. In this paper, we first clearly define these research questions and then design and perform a set of empirical experiments to address the questions. Finally, the experimental results are exhaustively discussed to answer the studied research questions. The empirical study and analysis presented in this paper could be informative for the management (modeling and forecasting) of the time-aware dynamic QoS attributes of cloud services. For example, we verify that machine-learning approaches are generally superior to the widely used statistical time-series methods in terms of both modeling accuracy and forecasting accuracy. Furthermore, after considering a variety of situations and cases, the GP-based approach is still the best option for the studied problem. In addition, except for the technical approaches, this paper also exhaustively studies the influence of the properties of the cloud dynamic QoS attributes, including their size and time granularity. [ABSTRACT FROM AUTHOR]

Aging sensitive parameters like dissipation factor, oil and paper conductivity and paper moisture can be estimated from insulation model parameterized using polarization current to assess the condition of transformer insulation. However, polarization current measurement is a time-consuming offline technique. During measurement, variation of environmental conditions (especially temperature) affects the monotonically decreasing nature of recorded data. Analysis of such affected data lead to incorrect conclusion regarding insulation condition. Power transformer being a crucial equipment, utility prefer to reduce its shut down time to minimum amount. In this paper a technique is discussed through which the polarization current measurement time can be reduced significantly. Several transformer data are used for verification of developed method. Presented results show that measurement data corresponding to only 10 minutes is sufficient to estimate the remaining data through the application of discussed method. [ABSTRACT FROM AUTHOR]

To maximize energy conversion, real-time control of a wave energy converter (WEC) requires knowledge of the present and future excitation force ($F_{\rm{ex}}$) acting on the device, which is a nonmeasurable quantity. The problem of estimation and forecasting of $F_{\rm{ex}}$ becomes more challenging when arrays of WECs are considered, due to the hydrodynamic interactions in the array. In this paper, a global $F_{\rm{ex}}$ estimator for a complete WEC array is developed and compared to a set of independent estimators, which utilize information local only to each device. A significant question is whether the array of measurements is sufficient to compensate for the greater complexity of the wave field, compared to the isolated body case. The paper shows that the global estimator is always more accurate than the independent estimator, improving up to $\text{45}\%$ the estimation accuracy of the independent estimator. Regarding prediction, two different $F_{\rm{ex}}$ forecasters for a WEC array are compared: a global forecaster, utilizing $F_{\rm{ex}}$ estimates from the full set of array devices, and an independent forecaster, utilizing only a local $F_{\rm{ex}}$ estimate. We demonstrate that the global forecaster achieves more accurate results, not only compared to the independent forecaster, but also compared to the isolated body case. [ABSTRACT FROM AUTHOR]

Multivariate time series (MTS) arise due to multisensor data collection in manufacturing. These data are complex in the sense that attributes have a varying scale, volitivity, continuity, and so on, and interattribute dependence also appears, which can mask the inherent information about system health status. Conventional machine learning-based process monitoring techniques are inefficient in predicting faults with MTS—their detection capability heavily relies on the input features, yet the classification power of MTS attributes is weakened by the complexity of multisensor data. Effective feature extraction is, therefore, necessary to facilitate fault prediction with MTS. This study proposes a fault prediction framework for MTS based on bag-of-words (BOW) feature extraction, statistical feature selection, and classification analysis. BOW models are for the first time adopted in a manufacturing context. Their superior capability in information preservation, local pattern recognition, and temporal effect accommodation has overcome the major limitations in current manufacturing practices with MTS. A comprehensive case study demonstrates the desirable performance of this framework on two MTS data sets from paper manufacturing and automotive manufacturing, as well as its superiority over conventional machine learning-based fault prediction. Note to Practitioners—Process monitoring in a multisensor environment has been a vital interest in manufacturing. The difficulty lies in the lack of detection power in many conventional techniques, e.g., control chart and logistic regression. A critical reason for such failure is the neglect of time effect in MTS—patterns associated with system faults tend to stretch a period, but a conventional control chart or classifier inspects each time stamp separately. Fault detection based on time series sequences is, therefore, essential. However, how to effectively extract features from MTS becomes a challenge. The framework proposed in this study adopts BOW models, specifically symbolic aggregate approximation (SAX), to extract features from MTS sequences, thus substantially improves the detection power against local patterns over time. Many manufacturing multisensor data are subject to such local patterns that point to the root cause of fault, so the proposed framework has a wide application in manufacturing. [ABSTRACT FROM AUTHOR]

Competition between Points of Interest (POIs) refers to the situation in which two POIs directly or indirectly provide similar services to secure businesses. A large portion of prior studies on competition analysis focuses on mining textual data, e.g., news articles and social comments. However, the increasing availability of human mobility and mobile query data enables a new paradigm for analyzing the competitive relationships among POIs, which remains largely unexplored. To this end, in this paper, we attempt to mine large-scale online map search query data for better understanding POI competitive relationships. Based on a co-query POI graph built from the map search query data, we develop a novel neural graphlet-based prediction framework to predict the competitive relationships among POIs. A unique perspective of our model is to infer latent POI competitive relationships by integrating multiple distinct factors, e.g., graphlet structure, geographical distance, and regional features, reflected in map search query data and POI data. Finally, we conduct extensive experiments on real-world datasets to demonstrate the effectiveness of the proposed framework, and show that our framework outperforms all baselines with a significant margin in all evaluation metrics. [ABSTRACT FROM AUTHOR]

An engineering change (EC) is an alteration made to a system that has been released following a system design process. EC propagation is a series of ECs occurring due to dependencies among components of a product. ECs can consume up to 50% of the overall engineering efforts during the development of a complex system. Therefore, EC propagation prediction received considerable attention in past decades as the product development industries started to suffer from the negative impacts of change propagation. This paper evaluates the current approaches to EC propagation prediction and presents a dynamic Bayesian networks (DBNs) approach to estimate change propagation risk (CPR) as well as a novel approach to estimate EC durations. Literature research shows that although some studies have used design structure matrices to estimate CPR and the total redesign duration (TRD) due to change propagation, an approach that allows iteration while accounting for the conjunction of all impacts has not been explored. This paper aims to fill the gaps for calculating CPR using DBN and evaluating change propagation paths from a Split-and task outcome logic, which accounts for the conjunction of all component relationships. This paper compares the proposed method results with the existing CPR and engineering change duration estimation methods using a real-world dataset from a U.S. Navy shipbuilding program. The results indicate that the CPR can be calculated using the proposed method without the shortcomings of the existing method and the accuracy for estimating engineering change durations is increased. [ABSTRACT FROM AUTHOR]

The paper proposes a novel block-wise prediction paradigm based on Convolutional Neural Networks (CNNs) for lossless video coding. A deep neural network model which follows a multi-resolution design is employed for block-wise prediction. Several contributions are proposed to improve neural network training. A first contribution proposes a novel loss function formulation for an efficient network training based on a new approach for patch selection. Another contribution consists in replacing all HEVC-based angular intra-prediction modes with a CNN-based intra-prediction method, where each angular prediction mode is complemented by a CNN-based prediction mode using a specifically trained model. Another contribution consists in an efficient adaptation of the CNN-based intra-prediction residual for lossless video coding. Experimental results on standard test sequences show that the proposed coding system outperforms the HEVC standard with an average bitrate improvement of around 5%. To our knowledge, the paper is the first to replace all the traditional HEVC-based angular intra-prediction modes with an intra-prediction method based on modern Machine Learning techniques for lossless video coding applications. [ABSTRACT FROM AUTHOR]

Short-term household electricity load forecasting is important for utility companies to ensure reliable power supplies. Traditional methods for load forecasting relied on historical records from one single data source and have limitations with insufficient or missing data. Recently, an emerging family of machine learning algorithms, multitask learning (MTL), has been developed and has the potential for load forecasting. By MTL, the electricity consumption data from multiple communities can be fused to improve forecasting accuracy. However, appropriate modeling of the relatedness to enable the between-community knowledge transfer remains a challenge. This paper proposes an improved MTL algorithm for a Bayesian spatiotemporal Gaussian process model (BSGP) to characterize the relatedness among the different residential communities. It hypothesizes on the similar impacts of environmental and traffic conditions on electricity consumption in improving short-term load forecasting. Furthermore, this paper proposes a low-ranked dirty model along with an iterative algorithm to improve the learning of model parameters under an MTL framework. This paper uses a real-world case study from two residential communities in Tallahassee, Fl, USA, to demonstrate the method effectiveness. The proposed method significantly outperforms state-of-the-art forecasting methods and effectively captures the relatedness to provide between-community knowledge transfer compared with other MTL methods. [ABSTRACT FROM AUTHOR]

This paper describes a video coding technology proposal submitted by MediaTek in April 2018 in response to a joint call for proposals (CfP) issued by ITU-T VCEG and ISO/IEC MPEG. The proposal uses the conventional block-based hybrid coding approach with a breakthrough design of block structure, substantial improvements of coding tools, and new inventions of loop filtering techniques. First, the block structure is designed using a highly flexible partitioning scheme, where each coding tree unit in a picture is split into multiple coding units (CUs) by a recursive quaternary tree followed by a recursive binary-ternary tree. The prediction unit (PU) and transform unit (TU) concepts are unified with the CU concept, i.e., PU and TU are basically the same as CU. Moreover, inter prediction, intra prediction, transform, quantization, in-loop filtering, and entropy coding are all improved. Finally, new loop filtering techniques are invented, where the proposed convolutional neural network loop filtering is of particular interest. The proposed video codec achieves 43.81% average bit rate savings based on objective measures compared to the High Efficiency Video Coding (HEVC) anchors and is among the best-performing CfP responses both objectively and subjectively. It has been partially adopted into the working draft of the Versatile Video Coding (VVC) standard. [ABSTRACT FROM AUTHOR]

Probabilistic load forecasting is gaining growing interest by researchers and practitioners. Multi-stage forecasting systems have recently demonstrated their effectiveness in increasing the overall performances. In this paper, we address the effect of pre-processing load time series using wavelet-based decompositions, before using quantile regression forests and random forests to build probabilistic forecasts. Four wavelet-based decompositions are specifically used for this task. Forecasts for the load components resulting from these transformations are obtained through distinct models, in order to increase the accuracy and to reduce the computational effort. Numerical applications based on the actual data published during the 2014 Global Energy Forecasting Competition are presented to evaluate the performance in a comparison with several benchmarks. [ABSTRACT FROM AUTHOR]

The demand for multitasking GPUs increases whenever the GPU may be shared by multiple applications, either spatially or temporally. This requires that GPUs can be preempted and switch context to a new application while already executing one. Unlike CPUs, context switching in GPUs is prohibitively expensive due to the large context states to swap out. There have been a number of efforts on reducing the overhead of preemption, through reducing the context sizes or overlapping context switching with execution. All those techniques are reactive approaches, meaning that context switching occurs when the preemption request arrives. In this paper, we propose a dynamic and proactive mechanism to reduce the latency of preemption. We observe that kernel execution is almost always preceded by known commands in both CUDA and OpenCL implementations. Hence, a preemption can be anticipated before the actual request arrives. We study such lead time and develop a prediction scheme to perform an early state saving. When the actual preemption is invoked, an incremental update relative to the previous saved state is performed, much like the conventional checkpointing mechanism. Our design can also choose to drain or checkpointing dynamically and accurately according to the feature of kernels in the runtime. This design effectively reduces the stall time of the preempting kernel due to context switching by 58.6%. Moreover, through careful handling of the saved state, we can also reduce the overall size of saved state by an average of 23.3%, compared with a full context switching. [ABSTRACT FROM AUTHOR]

Recently chaos theory has emerged as a powerful tool to address forecasting problems of nonlinear time series, since it is able to meet the dynamical and geometrical structures of very complex systems, reaching higher accuracy on the prediction values than the classical approaches. This paper aims at applying the chaos theory principles to different problems, in order to pursue high levels of accuracy on the predicted results. After the verification of the chaotic behavior of the datasets taken into analysis through the largest Lyapunov exponent research, the detection of the suitable embedding dimension and time delay has been carried out, in order to reconstruct the phase space of the underlying dynamical systems. Three different predictive methods have been proposed for different datasets. Finally, the performance comparison with the moving average model, a deep neural network based strategy, and a chaos theory based algorithm recently proposed in literature has been provided. [ABSTRACT FROM AUTHOR]

Load forecasting is one of the most important and studied topics in modern power systems. Most of the existing research on day-ahead load forecasting try to build a good model to improve the forecasting accuracy. The forecasted load is then used as the input to generation scheduling with the ultimate goal of minimizing the cost of generation schedules. However, existing day-ahead load forecasting models do not consider this ultimate goal at the training/forecasting stage. This paper proposes a task-based day-ahead load forecasting model labeled as LfEdNet that combines two individual layers in one model, including a load forecasting layer based on deep neural network (Lf layer) and a day-ahead stochastic economic dispatch (SED) layer (Ed layer). The training of LfEdNet aims to minimize the cost of the day-ahead SED in the Ed layer by updating the parameters of the Lf layer. Sequential quadratic programming (SQP) is used to solve the day-ahead SED in the Ed layer. The test results demonstrate that the forecasted results produced by LfEdNet can lead to lower cost of day-ahead SED at the expense of slight reduction in forecasting accuracy. [ABSTRACT FROM AUTHOR]

Recently, event organizers and researchers have advocated the development of novel technologies supporting crowd control, notably for public events. This paper presents a crowd monitoring system based on probe requests (PRs), which are Wi-Fi packets smartphones send periodically. By estimating the global rate at which nearby smartphones send PRs, Wi-Fi sensors can estimate crowd counts. The core contribution of this paper is a computationally tractable method that forecasts crowd counts up to thirty minutes in the future, with forecasts becoming available as soon as two hours of data are available. The forecasting method relies on autoregressive integrated moving average (ARIMA) models. Contributions also include two methods that compute prediction intervals associated with the forecasts, one of which is based upon generalized autoregressive conditional heteroskedasticity (GARCH) models. Recent real-world data from Winter Wonders 2018/2019 (an event that took place in Brussels, Belgium) notably demonstrate that the proposed forecasting method outperforms its immediate variations as well as baseline models (i.e., random walk models). [ABSTRACT FROM AUTHOR]

This paper presents a formulation to determine the appropriate power dispatch of an energy storage system, whose available energy is dependent on the charging/discharging pattern from previous time periods. The implementation structure is consistent with current dispatch algorithms used in microgrids, and the algorithm can be used in either grid-connected or islanded modes of operation. The proposed approach employs a backcasting algorithm to estimate the net stored energy value, against which the current cost of energy is compared to determine how the storage system should be used to perform arbitrage. The contribution of this paper is a means to include the time-dependent resource in traditional economic dispatch algorithms to reduce the cost of energy in a microgrid while enabling the arbitrage algorithm to continuously adapt to changing market conditions. Results show that the backcasting algorithm is able to reduce the average cost of energy by 8.14% and can reduce the average cost of energy by up to 72.3% of the ideal reduction, as determined by a perfect forecasting dispatch. [ABSTRACT FROM AUTHOR]

Maritime traffic prediction is critical for ocean transportation safety management. In this paper, we propose a novel knowledge assisted methodology for maritime traffic forecasting based on a vessel’s waterway pattern and motion behavior. The vessel’s waterway pattern is extracted through a proposed lattice-based DBSCAN algorithm that significantly reduces the problem scale, and its motion behavior is quantitatively modeled for the first time using kernel density estimation. The proposed methodology facilitates the knowledge extraction, storage, and retrieval, allowing for seamless knowledge transfer to support maritime traffic forecasting. By incorporating both the vessel’s waterway pattern and motion behavior knowledge, our solution suggests a set of probable coordinates with the corresponding probability as the forecasting output. The proposed forecasting algorithm is capable of accurately predicting maritime traffic 5, 30, and 60 min ahead, while its computation can be efficiently completed in milliseconds for single vessel prediction. Owing to such a high computational efficiency, an extensive predictive analysis of hundreds of vessels has been reported for the first time in this paper. A web-based prototype platform is implemented for Singapore waters to demonstrate the solution’s feasibility in a real-world maritime operation system. The proposed approaches can be generalized for other marine waters around the world. [ABSTRACT FROM PUBLISHER]

Real-time dynamic line rating forecasting can be classified into two critical steps of evaluating the operation risk, and avoiding the overload operation of transmission lines. The dynamic capacity of the transmission line is generally predicted for scheduling power networks in the day-ahead market. However, the forecasting error motivates researchers to propose real-time forecasting models to correct the day-ahead scheduling results based on the accurate data close to real-time values in the balancing market. Among the methods presented for real-time dynamic line rating forecasting, the single-point estimation type has been considered more than other forecasting types. Despite its safety, and efficiency, a single point estimation type suffers from several significant drawbacks. For instance, this forecasting type cannot indicate the probabilistic distribution function of forecasted value, which is used in novel balancing methods of market scheduling. The main purpose of this paper is to present a density forecast method to predict real-time dynamic line rating for covering the balancing market requirements. Dynamic stochastic general equilibrium (DSGE) is applied to achieve this aim. Due to incorporating stochastic volatility in the proposed real-time forecasting model, evaluating simulation results, and reference models highlight a significant improvement in the performance of the real-time density forecast of DLR. [ABSTRACT FROM AUTHOR]

This paper proposes an adaptive network fuzzy inference system (ANFIS) based decision support system that can provide European policy makers with systematic guidance in allocating and prioritizing scant public resources, whilst accomplishing the best growth performance at a regional level. We do so by taking the stance of the Smart Specialization Strategies, which aim at consolidating the regional strengths and make effective and efficient use of public investment in R&D. By applying the ANFIS method, we were able to understand how—and to what extent—the competitiveness drivers promoted technological development and how the latter contributes to the economic growth of European regions. We used socio–economic, spatial, and patent-based data to train, test, and validate the models. What emerges is that an increase of R&D investments enhances the regional employment rate and the number of patents per capita; in turn, by taking into account the several combinations of specialization and diversification indicators, this leads to an increase of the regional gross domestic product (GDP). [ABSTRACT FROM AUTHOR]

Video Anomaly detection in videos refers to the identification of events that do not conform to expected behavior. However, almost all existing methods cast this problem as the minimization of reconstruction errors of training data including only normal events, which may lead to self-reconstruction and cannot guarantee a larger reconstruction error for an abnormal event. In this paper, we propose to formulate the video anomaly detection problem within a regime of video prediction. We advocate that not all video prediction networks are suitable for video anomaly detection. Then, we introduce two principles for the design of a video prediction network for video anomaly detection. Based on them, we elaborately design a video prediction network with appearance and motion constraints for video anomaly detection. Further, to promote the generalization of the prediction-based video anomaly detection for novel scenes, we carefully investigate the usage of a meta learning within our framework, where our model can be fast adapted to a new testing scene with only a few starting frames. Extensive experiments on both a toy dataset and three real datasets validate the effectiveness of our method in terms of robustness to the uncertainty in normal events and the sensitivity to abnormal events. [ABSTRACT FROM AUTHOR]

When viewing omnidirectional images (ODIs), viewers can access different viewports via head movement (HM), which sequentially forms head trajectories in spatial-temporal domain. Thus, head trajectories play a key role in modeling human attention on ODIs. In this paper, we establish a large-scale dataset collecting 21,600 head trajectories on 1,080 ODIs. By mining our dataset, we find two important factors influencing head trajectories, i.e., temporal dependency and subject-specific variance. Accordingly, we propose a novel approach integrating hierarchical Bayesian inference into long short-term memory (LSTM) network for head trajectory prediction on ODIs, which is called HiBayes-LSTM. In HiBayes-LSTM, we develop a mechanism of Future Intention Estimation (FIE), which captures the temporal correlations from previous, current and estimated future information, for predicting viewport transition. Additionally, a training scheme called Hierarchical Bayesian inference (HBI) is developed for modeling inter-subject uncertainty in HiBayes-LSTM. For HBI, we introduce a joint Gaussian distribution in a hierarchy, to approximate the posterior distribution over network weights. By sampling subject-specific weights from the approximated posterior distribution, our HiBayes-LSTM approach can yield diverse viewport transition among different subjects and obtain multiple head trajectories. Extensive experiments validate that our HiBayes-LSTM approach significantly outperforms 9 state-of-the-art approaches for trajectory prediction on ODIs, and then it is successfully applied to predict saliency on ODIs. [ABSTRACT FROM AUTHOR]

Video saliency prediction (VSP) aims to imitate eye fixations of humans. However, the potential of this task has not been fully exploited since existing VSP methods only focus on modeling visual saliency of the input previous frames. In this paper, we present the first attempt to extend this task to video saliency forecasting (VSF) by forecasting attention regions of consecutive future frames. To tackle this problem, we propose a video saliency forecasting transformer (VSFT) network built on a new encoder-decoder architecture. Different from existing VSP methods, our VSFT is the first pure-transformer based architecture in the VSP field and is freed from the dependency of the pretrained S3D model. In VSFT, the attention mechanism is exploited to capture spatial-temporal dependencies between the input past frames and the target future frame. We propose cross-attention guidance blocks (CAGB) to aggregate multi-level representation features to provide sufficient guidance for forecasting. We conduct comprehensive experiments on two benchmark datasets, DHF1K and Hollywoods-2. We investigate the saliency forecasting and predicting abilities of existing VSP methods by modifying the supervision signals. Experimental results demonstrate that our method achieves superior performance on both VSF and VSP tasks. [ABSTRACT FROM AUTHOR]

Hospitals can predetermine the admission rate and facilitate resource allocation based on valid emergency requests and bed capacity estimation. The excess unoccupied beds can be determined with the help of forecasting the number of discharged patients. Extracting predictive features and mining the temporal patterns from historical observations are crucial for accurate and reliable forecasts. Machine learning algorithms have demonstrated the ability to learn temporal knowledge and make predictions for unseen inputs. This paper utilizes several machine learning algorithms to forecast the inpatient discharges of Singapore hospitals and compare them with statistical methods. A novel ensemble deep learning algorithm based on random vector functional links is established to predict inpatient discharges. The ensemble deep learning framework is optimized in a greedy layer-wise fashion. Several forecasting metrics and statistical tests are utilized to demonstrate the proposed method’s superiority. The proposed algorithm statistically outperforms the benchmark with a ranking of 1.875. Finally, practical implications and future directions are discussed. [ABSTRACT FROM AUTHOR]

The need of solar power uncertainty quantification in the power system has inspired probabilistic solar power forecasting. This paper proposes a novel multi-step parametric method for intra-day probabilistic solar power forecasting. First, statistical analysis on solar power distribution is done using four forecasting methods in real-world data. Fat tails are clearly found in solar power distribution, which could not be modelled by the widely-used normal distribution. In light of this discovery, two fat-tailed distributions, i.e., Laplace and two-sided power distributions, along with their generalized variants are then proposed to better model the conditional distribution of solar power output. Second, a recently proposed DeepAR model for time series probabilistic forecasting based on deep recurrent neural network is used to map various predictors into parameters of the fat-tailed distribution. Moreover, a novel loss function based on the continuous ranked probability score is proposed, and its closed-form formula over the proposed fat-tailed distributions is derived for efficient model training. Numerical results on public real-world data show that our method is very effective and the proposed model can provide intra-day probabilistic solar power forecasting with high quality and reliability. [ABSTRACT FROM AUTHOR]

Compared with commonly-used point forecasting, probabilistic forecasting provides quantitative information on the uncertainty associated with wind power output. However, most studies focus on modelling only a fixed set of quantiles or intervals, which cannot provide the entire information of wind power distribution. In this paper, we propose a non-parametric and flexible method for probabilistic wind power forecasting. First, the distribution of wind power output is specified by spline quantile function, which avoids assuming a parametric form and also provides flexible shape for wind power density. Then, autoregressive recurrent neural network is used to build the non-linear mapping from input features to the parameters of quadratic spline quantile function. A novel loss function based on continuous ranked probability score (CRPS) is designed to train the forecasting model. We also derive the closed-form solution of the integral required in computing the CRPS-based loss function in order to improve the computational efficiency in training. The effectiveness of our proposed method has been validated with public real-world data from Global Energy Forecasting Competition 2014. Compared with some advanced benchmarks, our proposed approach outperforms them at least six percent in terms of CRPS, indicating that it can provide probabilistic wind power forecasting with high quality. [ABSTRACT FROM AUTHOR]

The depth video contains large smooth contents with sharp edges. Since the deep learning-based color video orientated intra prediction methods pay no attention to the characteristics of depth video, they are unsuitable for optimizing the coding efficiency of depth video. In this paper, a multiple resolution prediction method with deep up-sampling is proposed to promote the coding efficiency of depth video. To efficiently encode the depth blocks of different complexity, the depth block is selectively encoded at different resolutions, including $\times 1$ , $\times 1$ /2, and $\times 1$ /4 resolutions. If the block is encoded with a low-resolution (LR), the resolution of reconstructed LR depth block is recovered by an up-sampling network. To constrain the quality of both reconstructed high-resolution depth block and its synthesized view, a view synthesis distortion guidance mechanism is proposed for the up-sampling network. In addition, a distillation-based lightweight up-sampling network is proposed to reduce the computational complexity. Experimental results demonstrate that the proposed multiple resolution prediction method obtains an average of 10.84% BD-rate saving in comparison with 3D-HEVC. [ABSTRACT FROM AUTHOR]

The rapid growth of mobile data traffic in a high-speed scenario imposes great challenges on wireless system design and optimization, where the inherent Doppler shift, frequent cell switching, and large penetration loss will significantly degrade the quality of transceiver signals. Wireless resource scheduling dynamically allocates wireless resources according to the user channel conditions and quality of service requirement, which is an enabling approach to maximize the resource utilization. In this paper, we develop a new wireless resource scheduling method in high mobility scenario to uplift the system performance. We first propose a feature extraction algorithm to extract the spatial traffic characteristics. On this basis, several time-series prediction algorithms are leveraged to predict the traffic volume. Then, we introduce cell switching indicator (CSI) into channel quality prediction to improve the signal-to-noise ratio (SNR) prediction performance. Based on the predicted results of service traffic and user channel quality variations, a resource scheduling strategy is proposed, which compromises the total system throughput and user satisfaction. Simulation results validate the effectiveness of the proposed scheduling algorithm. [ABSTRACT FROM AUTHOR]

The health status of the running gear in high-speed trains changes dynamically with time in a complete life cycle. Running gear systems composed of many coupled components are complex, and health statuses of which are difficult to predict in real time through a traditional health status prediction scheme. Lately, belief rule base (BRB), which is able to combine quantitative information and expert knowledge, has shown excellent expression in modeling complex systems. In the procedure of health status prediction, expert expertise can sufficiently enhance the accuracy and efficiency of this model. Therefore, this paper puts forwards a real-time health status prediction framework based on a multi-layer BRB with priority scheduling strategies for running gears. In the first-layer BRB, a time-series prediction model of multiple module BRB considering complete features is established. In the second-layer model, grey relation analysis (GRA) is employed in priority scheduling strategies of features. The third-layer BRB is used for assessing the health status of running gears by combining the features. In addition, the initial parameters of all module BRB given by experts may not be precise. Accordingly, the initial parameters in the BRB are updated by the recursive algorithm online. Finally, the proposed method is tested on the testing platform in running gears. The results make it clear the proposed method can predict the health status of running gears with much accuracy in real time. [ABSTRACT FROM AUTHOR]

The rapid increase in the penetration of renewable energy resources characterized by high variability and uncertainty is bringing new challenges to the power system operation. To ensure the efficient and reliable operation of electric grid, an accurate and general short-term forecasting algorithm with interpretability is desired. Moreover, the extensive off-site information provided by the proliferation of new renewable plants stimulates the interests in the spatiotemporal forecasting. In this paper, an attention temporal convolutional network, which is built on stacked dilated causal convolutional networks and attention mechanisms, is proposed to perform the ultra-short-term spatiotemporal forecasting of renewable resources. Compared with the existing spatiotemporal forecasting methods, the presented model needs no domain knowledge and can be applied to different forecasting tasks such as solar generation and wind speed forecasting. The attention mechanism improves the interpretability. The algorithm can be used to produce both point and probabilistic forecasts. Numerical results on the data sets from National Renewable Energy Laboratory show superior performance over five baselines, in terms of skill scores. Compared with the baselines, the average improvements of accuracy introduced by the proposed method for the point and probabilistic forecasting are 15.08% and 15.85%, respectively. [ABSTRACT FROM AUTHOR]

In recent decades, the evolution of loads and distributed energy resources has added great complexity and uncertainty to distribution networks. In such a context, a reliable characterization of the uncertainty associated with prediction is fundamental to the wide range of the newly emerging applications in the low-voltage level grid. In this regard, this paper proposes a novel method to solve the short-term probabilistic load forecasting (STPLF) problem in distribution networks in which the loads are usually too volatile to be forecasted accurately. The approach developed employs a Dirichlet process mixture model (DPMM) to handle the uncertainty of load patterns, which is inferred by a Markov chain Monte Carlo (MCMC)-based method. Thereafter, the DPMM representation of the load patterns is combined with a tree-based ensemble learning method to address the STPLF by solving a classification problem. The final result is averaged over all MCMC samples. The proposed technique is compared to selected benchmark methods at different aggregation levels using smart meter datasets collected from customers located in Ireland as well as the City of Saskatoon, SK, Canada. The results obtained demonstrate that the proposed approach outperforms the benchmark methods in STPLF at the given aggregation levels. [ABSTRACT FROM AUTHOR]

In this paper, we consider modeling and prediction of power loads due to fast charging stations for plug-in electric vehicles (EV). The first part of this paper is to simulate work of a fast charger activity by exploiting empirical data that characterize EV user behavior. The second part describes the time series obtained by this simulator and its properties. We show that the power load aggregated over a number of fast chargers (after deseasonalizing and elimination of the linear trend) is a self-similar process with the Hurst parameter 0.57

This paper presents a deep residual network for improving time-series forecasting models, indispensable to reliable and economical power grid operations, especially with high shares of renewable energy sources. Motivated by the potential performance degradation due to the overfitting of the prevailing stacked bidirectional long short-term memory (Bi-LSTM) layers associated with its linear stacking, we propose a concatenated residual learning by connecting the multi-level residual network (MRN) and DenseNet. This method further integrates long and short Bi-LSTM networks, ReLU, and SeLU for its activating function. Rigorous studies present superior prediction accuracy and parameter efficiency for the widely used temperature dataset as well as the actual wind power dataset. The peak value forecasting and generalization capability, along with the credible confidence range, demonstrate that the proposed model offers essential features of a time-series forecasting, enabling a general forecasting framework in grid operations. The source code of this paper can be found in https://github.com/MinseungKo/DRNet.git. [ABSTRACT FROM AUTHOR]

As we all know, science and technology are developing faster and faster. Many experts and scholars have demonstrated that human diseases are related to lncRNA, but only a few associations have been confirmed, and many unknown associations need to be found. In the process of finding associations, it takes a lot of time, so finding an efficient way to predict the associations between lncRNAs and diseases is particularly important. In this paper, we propose a multi-label fusion collaborative matrix factorization (MLFCMF) approach for predicting lncRNA-disease associations (LDAs). Firstly, the lncRNA space and disease space are optimized by multi-label to enhance the intrinsic link between lncRNA and disease and to tap potential information. Multi-label learning can encode a variety of data information from the sample space. Secondly, to learn multi-label information in the data space, the fusion method is used to handle the relationship between multiple labels. More comprehensive information will be obtained by weighing the effects of different labels. The addition of Gaussian interaction profile (GIP) kernel can increase the network similarity. Finally, the lncRNA-disease associations are predicted by the method of collaborative matrix factorization. The ten-fold cross-validation method is used to evaluate the MLFCMF method, and our method finally obtains an AUC value of 0.8612. Detailed analysis of ovarian cancer, colorectal cancer, and lung cancer in the simulation experiment results. So it can be seen that our method MLFCMF is an effective model for predicting lncRNA-disease associations. [ABSTRACT FROM AUTHOR]

Adaptive prediction is an important tool for efficient compression of non-stationary signals. A common approach to achieve adaptivity is to switch between a set of prediction modes, designed to capture variations in signal statistics. The design poses several challenges including: i) catastrophic instability due to statistical mismatch driven by propagation through the prediction loop, and ii) severe non-convexity of the cost surface that is often riddled with poor local minima. Motivated by these challenges, this paper presents a near-optimal method for designing prediction modes for adaptive compression. The proposed method builds on a stable, open-loop platform, but with a subterfuge that ensures that the design is asymptotically optimized for closed-loop operation. The non-convexity is handled by deterministic annealing, a powerful optimization tool to avoid poor local minima. To demonstrate the impact of the proposed approach on practical applications, we consider adaptive, transform-domain predictor design for enhancing standard video coding. Experimental results validate the benefits of the proposed design in terms of significant performance gains for both predictive compression systems in general and video coding in particular. [ABSTRACT FROM AUTHOR]