Data Decomposition Method Research Articles

Information Extraction from Satellite-Based Polarimetric SAR Data Using Simulated Annealing and SIRT Methods and GPU Processing

The main goal of this research was to propose a new method of polarimetric SAR data decomposition that will extract additional polarimetric information from the Synthetic Aperture Radar (SAR) images compared to other existing decomposition methods. Most of the current decomposition methods are based on scattering, covariance or coherence matrices describing the radar wave-scattering phenomenon represented in a single pixel of an SAR image. A lot of different decomposition methods have been proposed up to now, but the problem is still open since it has no unique solution. In this research, a new polarimetric decomposition method is proposed that is based on polarimetric signature matrices. Such matrices may be used to reveal hidden information about the image target. Since polarimetric signatures (size 18 × 9) are much larger than scattering (size 2 × 2), covariance (size 3 × 3 or 4 × 4) or coherence (size 3 × 3 or 4 × 4) matrices, it was essential to use appropriate computational tools to calculate the results of the proposed decomposition method within an acceptable time frame. In order to estimate the effectiveness of the presented method, the obtained results were compared with the outcomes of another method of decomposition (Arii decomposition). The conducted research showed that the proposed solution, compared with Arii decomposition, does not overestimate the volume-scattering component in built-up areas and clearly separates objects within the mixed-up areas, where both building, vegetation and surfaces occur.

Predicting behavior through dynamic modes in resting-state fMRI data

Dynamic properties of resting-state functional connectivity (FC) provide rich information on brain-behavior relationships. Dynamic mode decomposition (DMD) has been used as a method to characterize FC dynamics. However, it remains unclear whether dynamic modes (DMs), spatial-temporal coherent patterns computed by DMD, provide information about individual behavioral differences. This study established a methodological approach to predict individual differences in behavior using DMs. Furthermore, we investigated the contribution of DMs within each of seven specific frequency bands (0–0.1,...,0.6–0.7 Hz) for prediction. To validate our approach, we tested whether each of 59 behavioral measures could be predicted by performing multivariate pattern analysis on a Gram matrix, which was created using subject-specific DMs computed from resting-state functional magnetic resonance imaging (rs-fMRI) data of individuals. DMD successfully predicted behavior and outperformed temporal and spatial independent component analysis, which is the conventional data decomposition method for extracting spatial activity patterns. Most of the behavioral measures that were predicted with significant accuracy in a permutation test were related to cognition. We found that DMs within frequency bands <0.2 Hz primarily contributed to prediction and had spatial structures similar to several common resting-state networks. Our results indicate that DMD is efficient in extracting spatiotemporal features from rs-fMRI data.

Container Volume Prediction Using Time-Series Decomposition with a Long Short-Term Memory Models

The purpose of this study is to improve the prediction of container volumes in Busan ports by applying external variables and time-series data decomposition methods to deep learning prediction models. Previous studies on container volume forecasting were based on traditional statistical methodologies, such as ARIMA, SARIMA, and regression. However, these methods do not explain the complexity and variability of data caused by changes in the external environment, such as the global financial crisis and economic fluctuations. Deep learning can explore the inherent patterns of data and analyze the characteristics (time series, external environmental variables, and outliers); hence, the accuracy of deep learning-based volume prediction models is better than that of traditional models. However, this does not include the study of overall trends (upward, steady, or downward). In this study, a novel deep learning prediction model is proposed that combines prediction and trend identification of container volume. The proposed model explores external variables that are related to container volume, combining port volume time-series decomposition with external variables and deep learning-based multivariate long short-term memory (LSTM) prediction. The results indicate that the proposed model performs better than the traditional LSTM model and follows the trend simultaneously.

Decomposition-selection-ensemble forecasting system for energy futures price forecasting based on multi-objective version of chaos game optimization algorithm

Effective crude oil and natural gas futures price forecasting is a crucial endeavor for financial energy markets and is also a challenging work due to the nonlinear and fluctuant characteristic of futures price time series. Most existing researches have failed at the consideration of both linear and nonlinear information, optimal sub-model selection, and interval forecasting. To bridge these gaps, a novel decomposition-selection-ensemble forecasting system is proposed to perform deterministic prediction and interval forecasting in this study, which is constituted by data decomposition method, optimal sub-model selection strategy, proposed multi-objective version of chaos game algorithm, and multiple forecasting models. The proposed forecasting system prominently prompted the forecasting accuracy and stability of energy futures price, and improved the applicability at dealing with different data characteristic. Empirical results based on energy futures price demonstrated that the point forecasting and interval forecasting results obtained from the proposed forecasting system are more reliable and stable relative to other comparative models; thus, it can provide useful references for national economic policies and operators in financial energy markets.

A new approach for evaluating technology inequality and diffusion barriers: The concept of efficiency Gini coefficient and its application in Chinese provinces

The previous research has suggested that the existence of technology disparities reduces a contribution of environmental technological progress to combat various environmental issues. To handle the difficulty, we propose a new framework for assessing technology inequality and technology diffusion barriers. To measure them, we discuss the concept of efficiency Gini coefficient. Then, we combine it with Data Envelopment Analysis (DEA), DEA-DA (Discriminant Analysis) and decomposition methods of inequality measurement. Using the combined approach, we perform group-based and source-based decomposition to identify technology diffusion barriers. Empirically, this study has utilized the approach to evaluate technology inequality in Chinese provinces from 2007 to 2018. The main findings are summarized as follows. First, an overall efficiency Gini coefficient tended to increase over time, implying that technological disparity has increased across Chinese provinces. The disparity reduced the contribution of technological progress in addressing various environmental issues. Second, two factors mainly produced the technology inequality. They were cross-group inequality and within-group efficiency Gini coefficient. These results explained growing disparities across Chinese regions. Finally, a considerable difference existed in the within-group efficiency Gini coefficient.

Decomposition-based hybrid wind speed forecasting model using deep bidirectional LSTM networks

The goal of sustainable development can be attained by the efficient management of renewable energy resources. Wind energy is attracting attention worldwide due to its renewable and sustainable nature. Accurate wind speed prediction is essential for the stable functioning of wind turbines to generate wind power. However, the flexible and intermittent nature of wind speed makes accurate wind speed forecasting a challenging task. The proposed wind speed forecasting framework combines the features of various data decomposition techniques and Bidirectional Long Short Term Memory (BiDLSTM) networks. Presently, Data Decomposition models such as the Wavelet Transform are extensively employed for wind speed forecasting to improve the accuracy of the forecasting models. Hence, in this paper, various data decomposition techniques that can denoise the signal are investigated and applied to partition the input time series into several high and low-frequency signals. The data decomposition methods, namely, Wavelet Transform, Empirical Model Decomposition, Ensemble Empirical Mode Decomposition, and Empirical Wavelet Transform, have been applied to denoise the dataset. The low and high-frequency sub-series are forecasted separately using Bidirectional LSTM networks, and the forecasting outcomes of low and high-frequency signals are aggregated to get the final forecasting results. The empirical results establish that the proposed EWT- based hybrid model outperforms other decomposition-based models in accuracy and stability. The performance of the EWT-BiDLSTM model is further compared with Bidirectional LSTM networks with skip connections. The experimental results substantiate that the proposed decomposition-based hybrid deep BiDLSTM models with skip connections exhibit better prediction accuracy than other models.

Iterative Enhanced Multivariance Products Representation for Effective Compression of Hyperspectral Images

Effective compression of hyperspectral (HS) images is essential due to their large data volume. Since these images are high dimensional, processing them is also another challenging issue. In this work, an efficient lossy HS image compression method based on enhanced multivariance products representation (EMPR) is proposed. As an efficient data decomposition method, EMPR enables us to represent the given multidimensional data with lower-dimensional entities. EMPR, as a finite expansion with relevant approximations, can be acquired by truncating this expansion at certain levels. Thus, EMPR can be utilized as a highly effective lossy compression algorithm for hyper spectral images. In addition to these, an efficient variety of EMPR is also introduced in this article, in order to increase the compression efficiency. The results are benchmarked with several state-of-the-art lossy compression methods. It is observed that both higher peak signal-to-noise ratio values and improved classification accuracy are achieved from EMPR-based methods.

Short-term load forecasting model based on EEMD-LSTS-ARIMA

To improve the accuracy of the short-term power load forecasting, this paper designed a model, namely, the first to use based on empirical mode decomposition (EMD) method based on the collection of ensemble empirical mode decomposition (EEMD) in order to improve signal decomposition in the process of modal aliasing phenomenon and false component, and using the decomposition method of power load data decomposition, modal components with different characteristics, and according to its characteristics can be divided into high frequency components, intermediate frequency and low frequency components, Then the Long Short-Term Memory network (LSTM) is used for load forecasting. Using the autoregressive Integrated moving average model (Arima) to correct the predicted residual error, the load forecasting model based on EEMD-LSTM-ARIMA is established. Finally, the data of Queensland, Australia is used as the research sample. On the above prediction model, the data of this region is input into the model for power load prediction. At the end of the prediction, the final prediction results are compared with the EEMD-LSTM model through root mean square error (RMSE) and mean absolute percentage error (MAPE). Through comparison, it is found that the method used in this paper has a high prediction accuracy.

Spatial ensemble prediction of hourly PM2.5 concentrations around Beijing railway station in China

Air pollution has a serious negative impact on human health and economic development. Railway stations have the maximum flow of people in a city. It is necessary to provide precise air pollution forecasting for railway station areas. In this study, a spatial ensemble model is proposed to predict hourly PM2.5 concentrations for the Beijing railway station. In the proposed model, the spatial analysis is realized by a spatial feature selection and a spatial ensemble. The spatial feature selection method can recognize the correlated monitoring sites around the Beijing railway station and rank them. An optimization method is utilized to optimize the weighting coefficients of different correlated sites. Base predictors of different correlated sites are integrated according to the weighting coefficients. Besides, a data decomposition method is also utilized to enhance the performance of the spatial model. In this study, data processing methods and spatial analysis methods are combined with each other to build the spatial ensemble model. Four quarters of PM2.5 concentration data are utilized to verify the effectiveness and stability of the proposed model. The proposed spatial ensemble model can outperform other comparison models.

A Composite Uncertainty Forecasting Model for Unstable Time Series: Application of Wind Speed and Streamflow Forecasting

Wind speed and streamflow series always are nonlinear and unstable because the effects of chaotic weather systems. These inherent features make them difficult to forecast, especially in a changing environment. To improve forecasting accuracy, an innovation uncertainty forecasting architecture is developed by coupling data decomposition method, feature selection, multiple artificial intelligence (AI) techniques and composite strategy to do unstable time series forecasting. In the designed architecture, the AVMD (adaptive variational mode decomposition) is first applied to excavate implicit information from the original time series. Then, the random forest is utilized to select the suitable inputs for each mode. After that, the GPR (Gaussian Process Regression), a very famous probabilistic AI technique, is driven by various neural networks (ELM (Extreme Learning Machine), BP (Back Propagation Neural Networks), GRNN(Generalized Regression Neural Networks) and RBF (Radial Basis Function Neural Networks)) to produce both deterministic and probabilistic forecasting results in a nonlinear manner to play strengths of each other. The effectiveness and applicability of the proposed approach is verified by unstable wind speed data and streamflow data, and also compared with eleven related models. Results indicate that the proposed model not only improves the forecasting accuracy for deterministic predictions, but also provides more probabilistic information for decision making. The proposed method achieves significantly better performance than the traditional forecasting models both on wind speed forecasting and streamflow forecasting with at least 50% average performance promotion over all the eleven competitors. Comprehensive comparisons demonstrate the superior performance of the proposed method than the involved models as a powerful tool for unstable series forecasting.

Wind speed forecasting models based on data decomposition, feature selection and group method of data handling network

High precision wind speed forecasting plays an important role in the process of wind power conversion. In the study, a hybrid wind speed forecasting structure is proposed based on signal decomposition, feature selection and GMDH network. Several decomposition methods including the ensemble empirical mode decomposition, the wavelet packet decomposition, and two novel secondary decomposition method are utilized to decompose the original wind speed into several subseries, respectively. The binary-coded genetic algorithm is adopted to optimize the input features and arrange the input-output structure for the built predictors. The selected variables are put into the GMDH network to build the forecasting model for each subseries. According to the comparison results over 1-step to 6-step predictions, it can be concluded that the proposed models can get satisfactory results in wind speed forecasting.

Numbers in action during cognitive flexibility – A neurophysiological approach on numerical operations underlying task switching

Cognitive flexibility is a major competence to cope with daily life requirements and is usually investigated using task switching paradigms. Often, numerical stimuli are used to examine switching between task rules. Based upon functional neuroanatomical considerations we hypothesize that the ability to efficiently perform task switching varies depending on the cognitive operations performed with these numerical stimuli during task switching (magnitude vs parity judgments). We use a system-neurophysiological approach combining EEG and event-related potential (ERP) recordings with temporal data decomposition and source localization methods. We show that task switching processes are more demanding during parity judgments, compared to magnitude judgments. This, however, was only the case when task switching processes were triggered by external sensory stimuli, but not when memory-based processes had to be used during task switching. After accounting for intra-individual variability in the EEG data, the neurophysiological data showed that these effects were due to very specific subprocesses reflecting processes to update task sets and stimulus-response mappings during task switching. Source reconstructions show that left inferior and superior parietal areas (BA40 and BA7) were associated with these processes. The data show how different numerical operations differentially affect cognitive flexibility processes. Especially parity judgments exacerbate processes to update and reconfigure task sets during task switching in parietal areas. These findings are a valuable contribution to further reflections on the theories developed to date in task switching.

GPU-based lightweight parallel processing toolset for LiDAR data for terrain analysis

LiDAR products are provided at fine spatial resolutions and the data volume can be huge even for a small study region. Therefore, we have developed a parallel computing toolset that is built on Graphics Processing Units (GPUs) computing techniques to speed up the computational processes on LiDAR products. The toolset provides a set of fundamental processing functions for LiDAR point cloud data, serving as a basic toolkit to derive terrain data products. With this toolset, scientists with limited access to high-end computing facilities can still perform efficient analysis of LiDAR products without dealing with the technical complexity of developing and deploying tools for these products. We have integrated data decomposition methods to handle files that exceed the memory capacity of GPU devices. Preliminary results show that GPU-based implementation yields high speedup ratios and can handle files with a maximum size of 8 GB.

NVESTIGATION OF THE EFFICIENCY OF DISTRIBUTED INFORMATION SYSTEMS BASED ON THE PROCESSING OF LARGE AMOUNTS OF DATA

In this article investigated the features of processing large arrays of information for distributed systems. A method of singular data decomposition is used to reduce the amount of data processed, eliminating redundancy. Dependencies of com­putational efficiency on distributed systems were obtained using the MPI messa­ging protocol and MapReduce node interaction software model. Were analyzed the effici­ency of the application of each technology for the processing of different sizes of data: Non — distributed systems are inefficient for large volumes of information due to low computing performance. It is proposed to use distributed systems that use the method of singular data decomposition, which will reduce the amount of information processed. The study of systems using the MPI protocol and MapReduce model obtained the dependence of the duration calculations time on the number of processes, which testify to the expediency of using distributed computing when processing large data sets. It is also found that distributed systems using MapReduce model work much more efficiently than MPI, especially with large amounts of data. MPI makes it possible to perform calculations more efficiently for small amounts of information. When increased the data sets, advisable to use the Map Reduce model.

Dynamic Functional Magnetic Resonance Imaging Connectivity Tensor Decomposition: A New Approach to Analyze and Interpret Dynamic Brain Connectivity.

There is a growing interest in using so-called dynamic functional connectivity, as the conventional static brain connectivity models are being questioned. Brain network analyses yield complex network data that are difficult to analyze and interpret. To deal with the complex structures, decomposition/factorization techniques that simplify the data are often used. For dynamic network analyses, data simplification is of even greater importance, as dynamic connectivity analyses result in a time series of complex networks. A new challenge that must be faced when using these decomposition/factorization techniques is how to interpret the resulting connectivity patterns. Connectivity patterns resulting from decomposition analyses are often visualized as networks in brain space, in the same way that pairwise correlation networks are visualized. This elevates the risk of conflating connections between nodes that represent correlations between nodes' time series with connections between nodes that result from decomposition analyses. Moreover, dynamic connectivity data may be represented with three-dimensional or four-dimensional (4D) tensors and decomposition results require unique interpretations. Thus, the primary goal of this article is to (1) address the issues that must be considered when interpreting the connectivity patterns from decomposition techniques and (2) show how the data structure and decomposition method interact to affect this interpretation. The outcome of our analyses is summarized as follows. (1) The edge strength in decomposition connectivity patterns represents complex relationships not pairwise interactions between the nodes. (2) The structure of the data significantly alters the connectivity patterns, for example, 4D data result in connectivity patterns with higher regional connections. (3) Orthogonal decomposition methods outperform in feature reduction applications, whereas nonorthogonal decomposition methods are better for mechanistic interpretation.

A data decomposition method for stepwise migration of complex legacy data

SummarySooner or later, in almost every company, the maintenance and further development of large enterprise information technology (IT) applications reaches its limit. From the point of view of cost as well as technical capability, legacy applications must eventually be replaced by new enterprise IT applications. Data migration is an inevitable part of making this switch. While different data migration strategies can be applied, incremental data migration is one of the most popular strategies, due to its low level of risk: the entire data volume is split into several data tranches, which are then migrated in individual migration steps. The key to a successful migration is the strategy for decomposing the data into suitable tranches. This paper presents a method for data decomposition where the entire data volume of a monolithic enterprise IT application is split into independent data migration tranches. Each tranche comprises the data to be migrated in one migration step, which is usually executed during the application's downtime window. Unlike other methods, which describe data migration in a highly abstract way, we propose specific heuristics for data decomposition into independent data packages (tranches). The data migration approach described here is being applied in one of the largest migration projects currently underway in the European health care sector, comprising millions of customer records.

The study and application of a novel hybrid system for air quality early-warning

Air quality early-warning plays a vital role in improving air quality and human health, especially multi-step ahead air quality early-warning, which is significant for both citizens and environmental protection departments. However, most previous studies have only employed simple data decomposition to perform one-step forecasting and were aimed at enhancing forecasting accuracy or stability. Little research has improved these two standards simultaneously, leading to poor forecasting performance. Because of its significance, relevant research focused on multi-step ahead air quality early-warning is especially needed. Therefore, in this paper, a novel hybrid air quality early-warning system, which consists of four modules: data preprocessing module, optimization module, forecasting module and evaluation module, is proposed to perform multi-step ahead air quality early-warning. In this system, an effective data decomposition method called the modified complete ensemble empirical mode decomposition with adaptive noise is developed to effectively extract the characteristics of air quality data and to further improve the forecasting performance. Moreover, the hybrid Elman neural network model, optimized by the multi-objective salp swarm algorithm, is successfully developed in the forecasting module and simultaneously achieves high forecasting accuracy and stability. In addition, the evaluation module is designed to conduct a reasonable and scientific evaluation for this system. Three cities in China are employed to test the effectiveness of the proposed early-warning system, and the results reveal that the proposed early-warning system has superior ability in both accuracy and stability than other benchmark models and can be used as a reliable tool for multi-step ahead air quality early-warning.

A Review of Crude Oil Prices Forecasting using Hybrid Method

Crude oil is considered as a crucial energy source in modern days. Consequently, the fluctuation of crude oil prices can cause a significant impact on economic activities. Researchers have proposed many hybrid forecasting models on top of single forecasting methods which are utilized to predict crude oil prices movement more accurately. Nevertheless, many limitations still existed in hybrid forecasting models and models that can predict crude oil prices as accurate as possible is required. The motivations of this review paper are to identify and assess the mostly used crude oil prices forecasting methods and to analyse their current limitations. 12 studies that used “decomposition-and-ensemble” framework was selected for review. Wavelet transform is identified as the mostly used data decomposition method while some limitations have been recognized. Future researches should include more studies to further elucidate the limitations in existing forecasting method so that subsequent forecasting methods can be improved.

Data decomposition method for full-core Monte Carlo transport–burnup calculation

Monte Carlo transport simulations of a full-core reactor with a high-fidelity structure have been made possible by modern-day computing capabilities. Performing transport–burnup calculations of a full-core model typically includes millions of burnup areas requiring hundreds of gigabytes of memory for burnup-related tallies. This paper presents the study of a parallel computing method for full-core Monte Carlo transport–burnup calculations and the development of a thread-level data decomposition method. The proposed method decomposes tally accumulators into different threads and improves the parallel communication pattern and memory access efficiency. A typical pressurized water reactor burnup assembly along with the benchmark for evaluation and validation of reactor simulations model was used to test the proposed method. The result indicates that the method effectively reduces memory consumption and maintains high parallel efficiency.

Topic Modeling of Large Scale Social Text

In order to solve the problem of topic modeling for large-scale social short texts, this paper studied the parallel LDA modeling method and the dynamic topic model that can capture the dynamic characteristics of the topic. After that, a dynamic topic modeling method for large-scale text sets is proposed, which is based on the data decomposition and post-clustering method. It divided the whole corpus into independent fragments according to different features (e.g., time feature) and modeled the corpus in parallel. Then, it clustered the local topic in later stage. Experiments show that compared with DTM its execution time is less and it can capture the dynamic characteristics of the topic more effectively.