SVR Training Research Articles

Hybrid DE-Optimized GPR and NARX/SVR Models for Forecasting Gold Spot Prices: A Case Study of the Global Commodities Market

In this work, we highlight three different techniques for automatically constructing the dataset for a time-series study: the direct multi-step, the recursive multi-step, and the direct–recursive hybrid scheme. The nonlinear autoregressive with exogenous variable support vector regression (NARX SVR) and the Gaussian process regression (GPR), combined with the differential evolution (DE) for parameter tuning, are the two novel hybrid methods used in this study. The hyper-parameter settings used in the GPR and SVR training processes as part of this optimization technique DE significantly affect how accurate the regression is. The accuracy in the prediction of DE/GPR and DE/SVR, with or without NARX, is examined in this article using data on spot gold prices from the New York Commodities Exchange (COMEX) that have been made publicly available. According to RMSE statistics, the numerical results obtained demonstrate that NARX DE/SVR achieved the best results.

Research on methods for extracting aging characteristics and health status of lithium-ion batteries based on small samples

The number of aging samples is small with regard to the small sample characteristics of lithium-ion battery aging data, and the extraction of effective aging features is difficult to perform, which reduces the generalization of the battery aging behavior model as well as the accuracy of battery state of health (SOH) estimation. According to the incremental capacity analysis, a gray relational analysis (GRA) method based on PCA (principal component analysis) was proposed in this study, which was combined with AdaBoost-SVR (support vector regression) to achieve accurate SOH measurement under small sample conditions. Accordingly, the combination of GRA and PCA was able to fully mine the aging characteristic information from limited battery aging samples while improving the generalization ability of the battery aging behavior model. Moreover, based on AdaBoost-SVR, it was able to perform adaptive weighted sampling on small data samples, and through multiple iterations of the SVR training model, the full use of small sample information was realized, ensuring the accuracy of SOH estimation. Finally, by employing multiple sets of battery aging training sets and test sets for experimental verification, the joint algorithm based on GRA-PCA and AdaBoost-SVR was proven to achieve the accurate extraction of battery aging characteristics and precise tracking of capacity decline.

Feature Projection and SVR-Based Model Update for Object Tracking

Siamese network-based object tracking algorithms have recently gained popularity due to continuous improvement in tracking accuracy and robustness. However, these trackers are often limited by inefficient initialization and unstable online update. The model drift accumulates as more frames are processed. To address the above issues, this paper introduced a method that combines both deep learning and conventional regression algorithms (SVR, Support Vector Regression). Specifically, we propose a feature projection algorithm, which can effectively reduce the dimension of the features extracted from deep neural networks and improve the distinguishability of them at the same time. To make the features more robust for SVR training, we further propose two feature aggregation methods at both the channel level and the spatial level. We update the SVR model online instead of the deep neural network to make the tracking process more robust. Extensive experiments on four challenging benchmarks indicate that our proposed tracker is superior to baseline methods both qualitatively and quantitatively.

Power Plant Energy Predictions Based on Thermal Factors Using Ridge and Support Vector Regressor Algorithms

This work aims to model the combined cycle power plant (CCPP) using different algorithms. The algorithms used are Ridge, Linear regressor (LR), and upport vector regressor (SVR). The CCPP energy output data collected as a factor of thermal input variables, mainly exhaust vacuum, ambient temperature, relative humidity, and ambient pressure. Initially, the Ridge algorithm-based modeling is performed in detail, and then SVR-based LR, named as SVR (LR), SVR-based radial basis function—SVR (RBF), and SVR-based polynomial regression—SVR (Poly.) algorithms, are applied. Mean absolute error (MAE), R-squared (R2), median absolute error (MeAE), mean absolute percentage error (MAPE), and mean Poisson deviance (MPD) are assessed after their training and testing of each algorithm. From the modeling of energy output data, it is seen that SVR (RBF) is the most suitable in providing very close predictions compared to other algorithms. SVR (RBF) training R2 obtained is 0.98 while all others were 0.9–0.92. The testing predictions made by SVR (RBF), Ridge, and RidgeCV are nearly the same, i.e., R2 is 0.92. It is concluded that these algorithms are suitable for predicting sensitive output energy data of a CCPP depending on thermal input variables.

Condition Monitoring of DC-Link Electrolytic Capacitors in PWM Power Converters Using OBL Method

Since the lifespan of an electrolytic capacitor is relatively short compared to other power semiconductor devices, the failure rate accounts for 60% and, thus, it is the most vulnerable component of the power conversion device. Therefore, the accurate measurement of the lifetime of an electrolytic capacitor is very important in ensuring the reliability of the entire system, including the capacitor. In this paper, an online failure detection method for a DC-link electrolytic capacitor in a back-to-back Pulse width Modulation (PWM) converter using the opposition-based learning particle swarm optimization-based Support Vector Regression (OPSO-SVR) technique is proposed. In this method, the capacitance and the DC-link capacitor power have been used in offline mode for SVR training and testing. During the offline mode, the SVR parameters have been optimized with the OPSO algorithm to use online to estimate the real value of the DC-link capacitor. The experimental results prove the superiority of the proposed technique over the SVR.

Accurate disulfide-bonding network predictions improve ab initio structure prediction of cysteine-rich proteins.

Cysteine-rich proteins cover many important families in nature but there are currently no methods specifically designed for modeling the structure of these proteins. The accuracy of disulfide connectivity pattern prediction, particularly for the proteins of higher-order connections, e.g., >3 bonds, is too low to effectively assist structure assembly simulations. We propose a new hierarchical order reduction protocol called Cyscon for disulfide-bonding prediction. The most confident disulfide bonds are first identified and bonding prediction is then focused on the remaining cysteine residues based on SVR training. Compared with purely machine learning-based approaches, Cyscon improved the average accuracy of connectivity pattern prediction by 21.9%. For proteins with more than 5 disulfide bonds, Cyscon improved the accuracy by 585% on the benchmark set of PDBCYS. When applied to 158 non-redundant cysteine-rich proteins, Cyscon predictions helped increase (or decrease) the TM-score (or RMSD) of the ab initio QUARK modeling by 12.1% (or 14.4%). This result demonstrates a new avenue to improve the ab initio structure modeling for cysteine-rich proteins. http://www.csbio.sjtu.edu.cn/bioinf/Cyscon/ zhng@umich.edu or hbshen@sjtu.edu.cn. Supplementary data are available at Bioinformatics online.

Noise Removal using Support Vector Regression in Noisy Document Images

Noise removal of document images is a necessary step during preprocessing to recognize characters effectively because it has influences greatly on processing speed and performance for character recognition. We have considered using the spatial filters such as traditional mean filters and Gaussian filters, and wavelet transformed based methods for noise deduction in natural images. However, these methods are not effective for the noise removal of document images. In this paper, we present noise removal of document images using support vector regression. The proposed approach consists of two steps which are SVR training step and SVR test step. We construct an optimal prediction model using grid search with cross-validation in SVR training step, and then apply it to noisy images to remove noises in test step. We evaluate our SVR based method both quantitatively and qualitatively for noise removal in Korean, English and Chinese character documents, and compare it to some existing methods. Experimental results indicate that the proposed method is more effective and can get satisfactory removal results.

Online SVR Training by Solving the Primal Optimization Problem

Online estimation of regression functions becomes important in presence of drifts and rapid changes in the training data. In this article we propose a new online training algorithm for SVR, called Priona, which is based on the idea of computing approximate solutions to the primal optimization problem. For the solution of the primal SVR problem we investigated the trade-off between computation time and prediction accuracy for the gradient, diagonally scaled gradient, and Newton descent direction. The choice of a particular buffering strategy did not influence the performance of the algorithm. By using a line search Priona does not require a priori selection of a learning rate which facilitates its practical application. On various benchmark data sets Priona is shown to perform better in terms of prediction accuracy in comparison to the Norma and Silk online SVR algorithms. Further, tests on two artificial data sets show that the online SVR algorithms are able to track temporal changes and drifts of the regression function, if the buffer size and learning rate are selected appropriately.

Unified correlation for overall gas hold‐up in bubble column reactors for various gas–liquid systems using hybrid genetic algorithm‐support vector regression technique

AbstractThe objective of this study is to collect the data on overall gas hold‐up (∈G) for bubble column reactors handling various gas–liquid systems and further develop a unified data‐driven model for the estimation of the same. In this work, around 3300 experimental points for ∈G have been collected from 85 open sources spanning the years 1963–2008. The data‐driven model for overall gas hold‐up has been established using hybrid Genetic Algorithm‐Support Vector Regression (GA‐SVR)‐based methodology. In the present study, GA has been used for nonlinear rescaling of the parameters. These exponentially scaled parameters are subsequently subjected for SVR training. The technique is an extension of conventional SVR technique, showing relatively enhanced results. The proposed hybrid model is based on various prominent design and operating parameters (15 in number) which includes superficial gas velocity, superficial liquid velocity, gas density, molecular weight of gas, sparger type, sparger hole diameter, number of sparger holes, liquid viscosity, liquid density, liquid surface tension, ionic strength of liquid, operating temperature, operating pressure, liquid height, and the column diameter. The estimations made by the SVR‐based unified model for ∈G shows an excellent agreement with actual values with estimation accuracy of 98.5% and % AARE of 9.32%. For ease in applicability and ready reference of the practicing engineers, the hybrid GA‐SVR‐based model in the form of software and the entire database for ∈G has been uploaded on the link http://www.esnips.com/web/UICT‐NCL.

Development of Correlations for Overall Gas Hold-up, Volumetric Mass Transfer Coefficient, and Effective Interfacial Area in Bubble Column Reactors Using Hybrid Genetic Algorithm-Support Vector Regression Technique: Viscous Newtonian and Non-Newtonian Liquids

The objective of this study was to develop hybrid genetic algorithm−support vector regression (GA-SVR)-based correlations for overall gas hold-up (ϵG), volumetric mass-transfer coefficient (kLa), and effective interfacial area (a) in bubble column reactors for gas−liquid systems employing viscous Newtonian and non-Newtonian systems as the liquid phase. The hybrid GA-SVR is a novel technique based on the feature generation approach using genetic algorithm (GA). In the present study, GA has been used for nonlinear rescaling of attributes. These, exponentially scaled, are eventually subjected to SVR training. The technique is an extension of conventional SVR technique, showing relatively enhanced results. For this purpose an extensive literature search was done. From the published literature, 1629 data points for viscous Newtonian and 845 data points for viscous non-Newtonian systems for ϵG, 500 data points for viscous Newtonian and 556 data points for viscous non-Newtonian systems for kLa, and 208 data poin...