Support Vector Regression Method Research Articles

A Study on the Effect of Executive Control Network Functional Connection on the Therapeutic Efficacy of Repetitive Transcranial Magnetic Stimulation in Alzheimer's Disease.

Alzheimer's disease (AD) is a neurodegenerative disease characterized by brain network dysfunction. Few studies have investigated whether the functional connections between executive control networks (ECN) and other brain regions can predict the therapeutic effect of repetitive transcranial magnetic stimulation (rTMS). The purpose of this study is to examine the relationship between the functional connectivity (FC) within ECN networks and the efficacy of rTMS. We recruited AD patients for rTMS treatment. We established an ECN using baseline period fMRI data and conducted an analysis of the ECN's FC throughout the brain. Concurrently, the support vector regression (SVR) method was employed to project post-rTMS cognitive scores, utilizing the connectional attributes of the ECN as predictive markers. The average age of the patients was 66.86±8.44 years, with 8 males and 13 females. Significant improvement on most cognitive measures. We use ECN connectivity and brain region functions in baseline patients as features for SVR model training and fitting. The SVR model could demonstrate significant predictability for changes in Montreal Cognitive Assessment scores among AD patients after rTMS treatment. The brain regions that contributed most to the prediction of the model (the top 10% of weights) were located in the medial temporal lobe, middle temporal gyrus, frontal lobe, parietal lobe and occipital lobe. The stronger the antagonism between ECN and parieto-occipital lobe function, the better the prediction of cognitive improvement; the stronger the synergy between ECN and fronto-temporal lobe function, the better the prediction of cognitive improvement.

Robust clustering-based hybrid technique enabling reliable reservoir water quality prediction with uncertainty quantification and spatial analysis

Machine learning methodology has recently been considered a smart and reliable way to monitor water quality parameters in aquatic environments like reservoirs and lakes. This study employs both individual and hybrid-based techniques to boost the accuracy of dissolved oxygen (DO) and chlorophyll-a (Chl-a) predictions in the Wadi Dayqah Dam located in Oman. At first, an AAQ-RINKO device (CTD+ sensor) was used to collect water quality parameters from different locations and varying depths in the reservoir. Second, the dataset is segmented into homogeneous clusters based on DO and Chl-a parameters by leveraging an optimized K-means algorithm, facilitating precise estimations. Third, ten sophisticated variational-individual data-driven models, namely generalized regression neural network (GRNN), random forest (RF), gaussian process regression (GPR), decision tree (DT), least-squares boosting (LSB), bayesian ridge (BR), support vector regression (SVR), K-nearest neighbors (KNN), multilayer perceptron (MLP), and group method of data handling (GMDH) are employed to estimate DO and Chl-a concentrations. Fourth, to improve prediction accuracy, bayesian model averaging (BMA), entropy weighted (EW), and a new enhanced clustering-based hybrid technique called Entropy-ORNESS are employed to combine model outputs. The Entropy-ORNESS method incorporates a Genetic Algorithm (GA) to determine optimal weights and then combine them with EW weights. Finally, the inclusion of bootstrapping techniques introduces a stochastic assessment of model uncertainty, resulting in a robust estimator model. In the validation phase, the Entropy-ORNESS technique outperforms the independent models among the three fusion-based methods, yielding R2 values of 0.92 and 0.89 for DO and Chl-a clusters, respectively. The proposed hybrid-based methodology demonstrates reduced uncertainty compared to single data-driven models and two combination frameworks, with uncertainty levels of 0.24% and 1.16% for cluster 1 of DO and cluster 2 of Chl-a parameters. As a highlight point, the spatial analysis of DO and Chl-a concentrations exhibit similar pattern variations across varying depths of the dam according to the comparison of field measurements with the best hybrid technique, in which DO concentration values notably decrease during warmer seasons. These findings collectively underscore the potential of the upgraded weighted-based hybrid approach to provide more accurate estimations of DO and Chl-a concentrations in dynamic aquatic environments.

Inverse finite element method and support vector regression for automated crack detection with OFDR-Distributed fiber optic sensors

Crack detection is a crucial but challenging task. In this paper, a new method is introduced for automatical detection of the substrate cracks. An anomaly index is defined based on optical frequency domain reflectometry (OFDR) measured strain and inverse finite element method (iFEM) recovered strain. The index is sensitive to crack-induced strain but not sensitive to substrate strain. Leveraging on the self-defined anomaly index, the crack location can be identified with millimetre accuracy. Then, the detection of the crack depth is regarded as a typical multi-regression analysis and solved by support vector regression (SVR) method. Particle swarm optimization (PSO) approach is applied to find optimal parameters of the SVR kernel function. Experimental cases of a simple-supported beams with different crack locations and depths are performed, demonstrating the excellent prediction accuracy of the method. The proposed method has a promising potential in identification and quantification of cracks.

Inside Airbnb’s performance and adaptive strategies in Barcelona using artificial neural networks: A longitudinal, spatial, and multi-host perspective

This research explores the Airbnb platform's performance and adaptive strategies by analysing its spatial, temporal, and multi-host patterns. A three-layer model based on machine learning and neural networks, compared with a multiple linear regression, Random Forest Regression (RFR), and Support Vector Regression (SVR) methods, is used to conduct a longitudinal analysis of three representative months for tourism each year from 2016 to 2022. The study reveals the importance of “minimum nights”, active price management and professionalization, coupled with the potential transfer of accommodations in the medium- and long-term residential markets, as the platform's adaptive strategies. The findings also suggest a shift towards more professional host profiles and the consolidation of new tourist hubs in the city in post-Covid period. The study contributes to the understanding of Airbnb's performance and impact on global urban dynamics and demonstrates an application of machine learning to tourism and hospitality research. Theoretical and practical implications are discussed.

Elemental quantitation and evaluation of hydrocarbon in shale using fiber-optic laser induced breakdown spectroscopy

To meet the demands of in situ analysis, a fiber-optic laser-induced breakdown spectroscopy (FO-LIBS) system was established to predict the elemental concentrations and pyrolytic parameters. The key parameters, including pulse energy, gate delay, type of surrounding gas, and flow rate of gas, were optimized to improve the signal-to-noise ratio of the spectral lines, and the evolution characteristics of the plasma morphology in different gas atmospheres were investigated. Under the optimized experimental conditions, C, H, Si, Ca, Fe, and Mg were calibrated using pellet samples based on partial least squares regression, and the average relative prediction error of the natural samples was lower than 10%. In addition, the support vector regression method was utilized to predict pyrolytic parameters such as TOC, Pg, and Tmax with the R2 of all calibration curves being higher than 0.97. The abundance and generation potential of hydrocarbons were evaluated using the predicted TOC and Pg, and Tmax could contribute to the thermal maturity of the kerogen.

Reaction network design and hybrid modeling of S Zorb

At present, many countries are becoming more and more stringent in terms of sulfur content in fuel oil. S Zorb is a kind of desulfurization technology with advantages of exceptional desulfurization efficiency and small impact on octane number. To meet the needs of environmental requirements and the trend of digitalization in the petrochemical industry, a first-principle model of S Zorb was established based on industry data. In order to describe the desulfurization and the other side reactions, a reaction network was designed and the kinetic parameters were estimated by the particle swarm optimization algorithm. Two hybrid models based on the first-principle model and support vector regression method were established to correct the mass fraction of sulfur and predict the research octane number of the refined gasoline respectively. The results indicate that the hybrid models can predict the mass fraction of PIONA, sulfur content and research octane number of the refined gasoline accurately, of which the mean absolute percentage errors are less than 6%. Hybrid models were then applied to optimize the decision variables to minimize the research octane number loss. Optimization results show that the average reduction of the loss of research octane number is 21.8%, which suggests that the models developed hold promise for guiding practical production.

Fractional amplitude of low-frequency fluctuations in sensory-motor networks and limbic system as a potential predictor of treatment response in patients with schizophrenia

BackgroundPrevious investigations have revealed substantial differences in neuroimaging characteristics between healthy controls (HCs) and individuals diagnosed with schizophrenia (SCZ). However, we are not entirely sure how brain activity links to symptoms in schizophrenia, and there is a need for reliable brain imaging markers for treatment prediction. MethodsIn this longitudinal study, we examined 56 individuals diagnosed with 56 SCZ and 51 HCs. The SCZ patients underwent a three-month course of antipsychotic treatment. We employed resting-state functional magnetic resonance imaging (fMRI) along with fractional Amplitude of Low Frequency Fluctuations (fALFF) and support vector regression (SVR) methods for data acquisition and subsequent analysis. ResultsIn this study, we initially noted lower fALFF values in the right postcentral/precentral gyrus and left postcentral gyrus, coupled with higher fALFF values in the left hippocampus and right putamen in SCZ patients compared to the HCs at baseline. However, when comparing fALFF values in brain regions with abnormal baseline fALFF values for SCZ patients who completed the follow-up, no significant differences in fALFF values were observed after 3 months of treatment compared to baseline data. The fALFF values in the right postcentral/precentral gyrus and left postcentral gyrus, and the left postcentral gyrus were useful in predicting treatment effects. ConclusionOur findings suggest that reduced fALFF values in the sensory-motor networks and increased fALFF values in the limbic system may constitute distinctive neurobiological features in SCZ patients. These findings may serve as potential neuroimaging markers for the prognosis of SCZ patients.

Estimation of residual stress in dissimilar metals welding using deep fuzzy neural networks with rule-dropout

Welding processes are used to connect several components in nuclear power plants. These welding processes can induce residual stress in welding joints, which has been identified as a significant factor in primary water stress corrosion cracking. Consequently, the assessment of welding residual stress plays a crucial role in determining the structural integrity of welded joints. In this study, a deep fuzzy neural networks (DFNN) with a rule-dropout method, which is an artificial intelligence (AI) method, was used to predict the residual stress of dissimilar metal welding. ABAQUS, a finite element analysis program, was used as the data collection tool to develop the AI model, and 6300 data instances were collected under 150 analysis conditions. A rule-dropout method and genetic algorithm were used to optimize the estimation performance of the DFNN model. DFNN with the rule-dropout model was compared to a deep neural network method, known as a general deep learning method, to evaluate the estimation performance of DFNN. In addition, a fuzzy neural network method and a cascaded support vector regression method conducted in previous studies were compared. Consequently, the estimation performance of the DFNN with the rule-dropout model was better than those of the comparison methods. The welding residual stress estimation results of this study are expected to contribute to the evaluation of the structural integrity of welded joints.

Design and optimization of haze prediction model based on particle swarm optimization algorithm and graphics processor

With the rapid expansion of industrialization and urbanization, fine Particulate Matter (PM2.5) pollution has escalated into a major global environmental crisis. This pollution severely affects human health and ecosystem stability. Accurately predicting PM2.5 levels is essential. However, air quality forecasting currently faces challenges in processing vast data and enhancing model accuracy. Deep learning models are widely applied for their superior learning and fitting abilities in haze prediction. Yet, they are limited by optimization challenges, long training periods, high data quality needs, and a tendency towards overfitting. Furthermore, the complex internal structures and mechanisms of these models complicate the understanding of haze formation. In contrast, traditional Support Vector Regression (SVR) methods perform well with complex non-linear data but struggle with increased data volumes. To address this, we developed CUDA-based code to optimize SVR algorithm efficiency. We also combined SVR with Genetic Algorithms (GA), Sparrow Search Algorithm (SSA), and Particle Swarm Optimization (PSO) to identify the optimal haze prediction model. Our results demonstrate that the model combining intelligent algorithms with Central Processing Unit-raphics Processing Unit (CPU-GPU) heterogeneous parallel computing significantly outpaces the PSO-SVR model in training speed. It achieves a computation time that is 6.21–35.34 times faster. Compared to other models, the Particle Swarm Optimization-Central Processing Unit-Graphics Processing Unit-Support Vector Regression (PSO-CPU-GPU-SVR) model stands out in haze prediction, offering substantial speed improvements and enhanced stability and reliability while maintaining high accuracy. This breakthrough not only advances the efficiency and accuracy of haze prediction but also provides valuable insights for real-time air quality monitoring and decision-making.

Genetic Algorithms Application for Pricing Optimization in Commodity Markets

The perishable nature of vegetable commodities poses challenges for superstores, as reselling them is often unfeasible due to their short freshness period. Reliable market demand analysis is crucial for boosting revenue. This study simplifies the pricing and replenishment decision-making process by making reasonable assumptions about the selling time, wastage rate, and replenishment time for vegetable commodities. A single-objective planning model with the objective of profit maximization was constructed by fitting historical data using the nonparametric method of support vector regression (SVR). The study reveals a specific relationship between sales volume and cost-plus pricing for each category and predicts future cost changes using an LSTM model. Combining these findings, we substitute the relationship between sales volume and pricing as well as the LSTM prediction data into the model, and solve it using genetic algorithms in machine learning to derive the optimal replenishment volume and pricing strategy. Practical results show that the method can provide reasonable pricing and replenishment strategies for vegetable superstores, and after careful accounting, we arrive at an expected profit of RMB 22,703.14. The actual profit of the supermarket was RMB 19,732.89. The method, therefore, increases the profit of the vegetable superstore by 13.08%. By optimizing inventory management and pricing decisions, the superstore can better meet the challenges of vegetable commodities and achieve sustainable development.

Analysis of Damage Assessment Tweets During Disaster using Sentiment Analysis

This seems to be an abstract or summary of a paper on monitoring Twitter for damage assessments after a disaster. Using simple linear regression and Support Vector Regression methods for weighting and the random forest methodology for classification, the research provides a novel approach that makes use of low-level lexical characteristics, top-most frequency word features, and syntactic elements relevant to damage assessment. The accuracy of the suggested method for identifying damage assessment tweets is 94.62%, as measured across 14 typical disaster datasets for binary and multi-class categorization. Significant advancements were observed when comparing the proposed method to the state-of-the-art for both in-domain and cross-domain scenarios. The suggested method does not require labelled tweets or tweets of a specific disaster kind in order to be trained and implemented; instead, it can be trained on historical disaster datasets.

A horizontal vibration prediction method of high-speed elevator based on transferred-digital twin model

The vibrations of high-speed-elevators (HSEs) significantly impact the comfort of elevator rides, with horizontal vibrations (HV) being the most sensitive to bodies. Customized elevator products are characterized by high-speed capabilities, customization options, and complex operational conditions. These characteristics often lead to challenges such as difficulties in predicting HV, low prediction accuracy, and limited reusability of prediction models and data. In order to mitigate HV in elevators, its crucial to implement HV prediction during the design phase. To address this need, this paper proposes a method for predicting HSEHV based on a Transferred Digital Twin (TLDT) model. The HSE’s design is physically extracted, and an elevator Digital Twin (DT) model with a digital layer is constructed. This model includes product geometry models, dynamics models, and simulation models. A transfer learning (TL) model is developed to integrate simulation data generated by the digital layer with measurement data acquired from the data interface layer, resulting in high-capacity and high-fidelity DT data. Using the support vector regression (SVR) method, a high-dimensional nonlinear HSEHV prediction model is established, and is trained and optimized using DT data to achieve data-driven HSEHV prediction. The validity of this approach is confirmed using measured data from an elevator test tower. Results indicate that the SVR-based TLDT model achieves the highest accuracy and the lowest absolute error values for both peak-to-peak and A95 vibration acceleration when compared to other models such as NN, KNN, decision tree, and lasso models. Furthermore, by incorporating DT data, the model accuracies improve by 16.9%, 14.5%, 8.6%, 14.1%, and 8.6%, and 12.3%, 12.4%, 19.7%, 7.6%, and 8.7%, respectively, for the mentioned models. Finally, the HV of KLK2 elevator car system is predicted. When compared with simulation analysis and the conventional SVR prediction model, obtained results through the proposed method closely align with the measured values.

Application of the Support Vector Regression Method with the Grid Search Algorithm to Predict Movement Gold Price

Gold is an investment with the smallest risk because it can be sold anytime and anywhere. In Indonesia, gold bullion as an investment product is known for its purity level of 99.99%, namely gold bullion produced by PT. Aneka Tambang (Antam) through its Precious Metals business unit. Apart from its pure production, Antam gold bullion is easier to resell anytime and anywhere because it has an official certificate from the international gold standardisation institution, namely LBMA (London Bullion Market Association), to more easily estimate the value of gold bullion when sold. To overcome this, predictions of future gold prices are needed. In this research, one of the prediction methods is Support Vector Regression with the Grid Search Algorithm. In this method this method will be used to predict the price of gold, which aims to predict and find out the price of gold one year in the future to produce a level accuracy (MAPE) of 5.43% and the prediction of gold prices increasing from 2023-June-01 to 2024-March-23 while experiencing a decline starting in 2024-March-24. Research by examining the relationship between variables, which emphasises data consisting of numbers so that it is analysed based on statistical procedures using the Support Vector Regression method with data sourced from the daily price of gold bullion through PT. Gallery 24 Pawnshops, North Sumatra. Where this method is very well used in predicting by choosing the best kernel used is the linear kernel because, from these three kernels, the best hyperparameters were obtained for predicting gold price movements using a linear kernel with a division for training and testing data of 60: 40. The MAPE value obtained was 5.43.

Adaptive locally weighted support vector algorithm with asymmetrically parametric insensitive/margin model

In many fields such as finance and electrical load, locally weighted support vector regression (LWSVR) method can achieve better prediction results by constructing different sub-models for each prediction target. However, it selected a fixed and single nearest neighbor parameter of kglobal for each query, and adopted the traditional k-nearest neighbor (KNN) algorithm, which is suboptimal. The weights in LWSVR assigned to the neighboring sample points also affected the accuracy of the model predictions. In this paper, we propose an adaptive locally weighted support vector regression with asymmetrically parametric insensitive/margin model, called Asy-par-v-ALWSVR. Since the selection of similar sample for query points is significant. We propose a novel locally q-neighbour selection algorithm that adaptively selects the most suitable number of neighbours for each query point based on Mahalanobis distance, rather than using a fixed kglobal. This significantly improves prediction accuracy. The weights of neighbours for each query point in Asy-par-v-ALWSVR is based on the distance within the kernel-induced feature space. This effectively discerns the importance of samples during the training process. In particular, Asy-par-v-ALWSVR can handle skewed noise with heteroscedastic structure in regression problems. Finally, we evaluate the proposed model on a synthetic sinc dataset and 14 real datasets. The results demonstrate that Asy-par-v-ALWSVR outperforms six state-of-the-art SVR-based models.

Analysis of Spectral Characteristics of Cotton Leaves at Bud Stage under Different Nitrogen Application Rates

Soil salinity affects nutrient uptake by cotton. The cotton bud stage is a very important period in the process of cotton planting and directly affects the yield of cotton. The nutritional status of the bud stage directly affects the reflectance spectra of cotton canopy leaves. Therefore, it is of great significance to nondestructively monitor the nutritional status of the cotton bud stage on salinized soil via spectroscopic techniques and perform corresponding management measures to improve cotton yield. In this study, potted plants with different nitrogen application rates were set up to obtain the reflection spectral curves of cotton bud stage leaves, analyze their spectral characteristics under different nitrogen application rates, and establish spectral estimation models of chlorophyll density. The results are as follows: in the continuum removal spectrum of the cotton bud stage, the lowest point of the absorption valley near 500 nm shifted to the shortwave direction with an increasing nitrogen application rate. The mean reflectance between 765 and 880 nm was significantly different between nitrogen-stressed and nitrogen-unstressed cotton. The average reflectance of the near-infrared band, the absorption valley depths near 500 nm and 675 nm, the first derivative of the 710 nm reflectance, and the second derivatives of the 690 nm and 730 nm reflectance increased with increasing nitrogen application and chlorophyll density, and significant correlations were observed with the chlorophyll density. These parameters were modeled using support vector regression (SVR) and artificial neural network (ANN) methods, two commonly used algorithms in the field of machine learning. The determination coefficients of the three chlorophyll samples via the ANN models were 0.92, 0.77, and 0.94 for the modeling set and 0.77, 0.69, and 0.77 for the verification set. The ratio of quartile to root-mean-square error (RPIQ) of the ANN model was greater than 2.2, and the ratio of the standard error of the measured value to the standard error of the predicted (SEL/SEP) was close to 1, indicating that the chlorophyll density estimation models built based on the ANN algorithm had robust prediction ability. Our model could accurately estimate the leaf chlorophyll density in the cotton bud stage.

Augmented support vector regression with an autoregressive process via an iterative procedure

The Support Vector Regression (SVR) technique can approximate intricate systems by addressing learning and estimation challenges within a reproducing kernel Hilbert space, devoid of reliance on specific parameter assumptions. However, when dealing with correlated data like time series, the SVR method often falls short in accounting for underlying temporal structures, leading to limited enhancements in prediction efficiency. We introduce an enhanced SVR method that considers temporal correlations (TemporalSVR) to overcome this constraint. Our proposed method extends kernel functions to include additional linear kernels, facilitating learning temporal patterns. Additionally, we develope an iterative training procedure for the augmented regression model. During model training, we estimate the hyper-parameter in the corresponding loss function using a ‘working’ likelihood approach, enhancing the generalization capabilities of the proposed regression. To demonstrate superior forecasting performance, we conduct extensive numerical simulations on both linear and nonlinear systems and the TemporalSVR achieves improvements ranging from 8% to 114% based on the RMSE ratio from the AR-X model. Furthermore, we investigate the forecasting performance of three basic models (NARX-NN, Statistical SVR, and SVR-ARIMA) and four deep learning (DL) techniques (Transformer, Informer, Reformer, Autormer, and Autoformer) by using a WTI forecasting study. Our proposed TemporalSVR achieves the smallest RMSE at 2.22 and attains the highest success ratio of stock direction prediction at 71.30%. All these numerical results highlight the effectiveness and advantages of our TemporalSVR in handling temporal data and making accurate predictions.

Evaluation of the performance of SVR and ANN Models in Sea Water Level Prediction for the Izmir Coast of the Aegean Sea

Prediction of sea water level is a very important phenomenon for making future projections for flood control, human living conditions and coastal planning. However, it is not easy to estimate sea water level due to the influence of atmospheric conditions. Thus, Artificial Neural Networks (ANN) and Support Vector Regression (SVR) methods are used for the prediction of seawater level of Izmir Coast based on the time series data. Coefficient of determination (R2 ) and root mean square error (RMSE) are applied as model evaluation criteria in this study. 19 months of seawater level data in the time series were used in this study. The results show that the ANN model can predict the water level with R2 of 0.84 and 0.68 for 1st and 2nd days, respectively, while the SVR model can predict the water level with R2 of 0.83 and 0.69 for 1st and 2nd days, respectively.

Comparison of ANN and SVR based models in sea level prediction for the Black Sea coast of Sinop

Seawater level oscillations are very critical to coastal construction, flood prevention and human living conditions. However, it is difficult to accurately project the daily future for seawater level due to the effects of wind, precipitation and other atmospheric conditions. For this reason, in this paper, artificial intelligence (AI) based Artificial Neural Networks (ANN) and Support Vector Regression (SVR) methods are applied for the estimation of seawater level in Sinop Coast. In addition, Multiple Linear Regression (MLR) is used as a benchmarking model. In this study, coefficient of determination (R2) and root mean square error (RMSE) were applied as model evaluation criteria. Besides, 15 minutes (approximately 22 months) sea water level data of Sinop Station were collected and used as is. The findings revealed that the ANN model can predict the water level for 1st, 2nd, 3rd, 4th days with correlation coefficients (R2) of 0.84, 0.67, 0.64, 0.63, respectively, and the SVR model can predict for 1st, 2nd days with correlation coefficients (R2) of 0.86, 0.66, respectively.

SPATIAL PREDICTION OF RECEIVED SIGNAL STRENGTH FOR CELLULAR COMMUNICATION USING SUPPORT VECTOR MACHINE AND K-NEAREST NEIGHBOURS REGRESSION

Abstract. Signal strength maps are of great importance for cellular system providers in network planning and operation. Accurate prediction of signal strength is important for solving problems such as link quality. In this study, Received Signal Strength (RSS) prediction model is proposed for the 900 MHz band in the Van Yüzüncü Yıl University campus environment by using machine learning regression methods such as K- Nearest Neıghbours (KNN) and Support Vector Regression (SVR) together with Geographic Information Systems. For the training of this model, signal strength values taken from the RF Spectrum Analyser at different locations and distances were used. In addition, spatial data sets such as the digital elevation model, location of base stations and measurement stations, building heights and location, and land use/cover were used in the model. The effect of these data sets on RSS power is included in the model. The model aims to predict RSS accurately, visualize the estimated signal strength, and analyze the signal field strength coverage. Different kernels from the SVR model such as Polynomial, , and Sigmoid were tested. To increase the success of the model, appropriate parameter values were selected and configured according to SVR and KNN methods. For 900 MHz, the performances of SVR and KNN models were compared and the results of the models were verified using root mean squares (RMSE). Among the measured data, the lowest prediction is found in KNN Manhattan. According to the results of the simulation was observed that the SVR model created with spatial data performs better for Signal Strength. Finally, the lowest RMSE value (1.71 dB) was obtained from the Sigmoid kernel in the best signal strength estimation SVR model. The SVR model is recommended for Campus Area signal strength estimation.

Forecasting peak electric load: Robust support vector regression with smooth nonconvex ϵ‐insensitive loss

AbstractPeak power load forecasting is a key part of the commercial operation of the energy industry. Although various load forecasting methods and technologies have been put forward and tested in practice, the growing subject of tolerance for abnormal accidents is to develop robust peak load forecasting models. In this paper, we propose a robust smooth non‐convex support vector regression method, which improves the robustness of the model by adjusting adaptive control loss values and adaptive robust parameters and by reducing the negative impact of outliers or noise on the decision function. A concave‐convex programming algorithm is used to solve the non‐convexity of the optimization problem. Good results are obtained in both linear regression model and nonlinear regression model and two real data sets. An experiment is carried out in a power company in Jiangxi Province, China, to evaluate the performance of the robust smooth non‐convex support vector regression model. The results show that the proposed method is superior to support vector regression and generalized quadratic non‐convex support vector regression in robustness and generalization ability.