Least squares support vector machine with self-organizing multiple kernel learning and sparsity

Liu, Chang; Tang, Lixin; Liu, Jiyin

doi:10.1016/j.neucom.2018.11.067

Cited by 32 publications

(7 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Generally, in the SVM model, kernels such as radial basis, linear, sigmoid, and polynomial have been utilized by existing studies. Because of the below-optimal performance of these kernels for many tasks, multiple customized kernels have also been used by several studies [41][42][43]. This work combines kernel functions to form a customized kernel using Mercer's theorem [44].…”

Section: Prediction Of Students' Performancementioning

confidence: 99%

See 1 more Smart Citation

Predicting Students’ Academic Performance with Conditional Generative Adversarial Network and Deep SVM

Sarwat

Ullah

Sadiq

et al. 2022

Sensors

View full text Add to dashboard Cite

The availability of educational data obtained by technology-assisted learning platforms can potentially be used to mine student behavior in order to address their problems and enhance the learning process. Educational data mining provides insights for professionals to make appropriate decisions. Learning platforms complement traditional learning environments and provide an opportunity to analyze students’ performance, thus mitigating the probability of student failures. Predicting students’ academic performance has become an important research area to take timely corrective actions, thereby increasing the efficacy of education systems. This study proposes an improved conditional generative adversarial network (CGAN) in combination with a deep-layer-based support vector machine (SVM) to predict students’ performance through school and home tutoring. Students’ educational datasets are predominantly small in size; to handle this problem, synthetic data samples are generated by an improved CGAN. To prove its effectiveness, results are compared with and without applying CGAN. Results indicate that school and home tutoring combined have a positive impact on students’ performance when the model is trained after applying CGAN. For an extensive evaluation of deep SVM, multiple kernel-based approaches are investigated, including radial, linear, sigmoid, and polynomial functions, and their performance is analyzed. The proposed improved CGAN coupled with deep SVM outperforms in terms of sensitivity, specificity, and area under the curve when compared with solutions from the existing literature.

show abstract

Section: Prediction Of Students' Performancementioning

confidence: 99%

“…Existing studies show that the choice of the kernel for SVM has a substantial impact on its performance [41][42][43]. In addition, modified and custom kernels have been utilized for better results.…”

Section: Selection Of Appropriate Kernel Typementioning

confidence: 99%

Predicting Students’ Academic Performance with Conditional Generative Adversarial Network and Deep SVM

Sarwat

Ullah

Sadiq

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…The major shortcoming of SVM is the computational burden for optimization programming constrained. This weakness has been overcomed in Least-Square Support Vector Machines (LSSVM), which substitute a quadratic programming with linear equations [15][16][17][18]. The accuracy of the prediction of LSSVM depends on the hyperparameters value setting.…”

Section: Introductionmentioning

confidence: 99%

Long-term load forecasting using grey wolf optimizer -least-squares support vector machine

Yasin

Salim

Aziz

et al. 2020

IJ-AI

View full text Add to dashboard Cite

<p><span lang="EN-US">Long term load forecasting data is important for grid expansion and power system operation. Besides, it also important to ensure the generation capacity meet electricity demand at all times. In this paper, Least-Square Support Vector Machine (LSSVM) is used to predict the long-term load demand. Four inputs are considered which are peak load demand, ambient temperature, humidity and wind speed. Total load demand is set as the output of prediction in LSSVM. In order to improve the accuracy of the LSSVM, Grey Wolf Optimizer (GWO) is hybridized to obtain the optimal parameters of LSSVM namely GWO-LSSVM. Mean Absolute Percentage Error (MAPE) is used as the quantify measurement of the prediction model. The objective of the optimization is to minimize the value of MAPE. The performance of GWO-LSSVM is compared with other methods such as LSSVM and Ant Lion Optimizer – Least-Square Support Vector Machine (ALO-LSSVM). From the results obtained, it can be concluded that GWO-LSSVM provide lower MAPE value which is 0.13% as compared to other methods.</span></p>

show abstract

“…However, the selection of hyperparameters is still an open issue at present. Although these parameters can be obtained by optimization algorithms such as genetic algorithm and difference evolution [17,18,19], they are time consuming and obtain only suboptimal solutions [20]. Therefore, the above researches lack an adaptive selection mechanism for hyperparameters.…”

Section: Introductionmentioning

confidence: 99%

Multisource Latent Feature Selective Ensemble Modeling Approach for Small-Sample High-Dimensional Process Data in Applications

Zhang

Yu²,

Zhang³

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Several difficult-to-measure production qualities or environment pollution indices of industrial process must be measured using offline laboratory instruments. Soft measurement method is often used to perform online prediction of such parameters. Only small-sample modeling data with high-dimensional input features can be obtained due to the limitations and complex characteristics of the measurement device and process, respectively. Therefore, a new multisource latent feature selective ensemble (SEN) modeling approach is proposed in this study. First, input features are divided into different subgroups according to the characteristics of the modeling data. Second, the extracted multisource latent features evolve from the multi-layered selection algorithms, which are specified by feature reduction ratio, feature contribution ratio and mutual information value orderly for each subgroup. Finally, in order to construct candidate sub-models, an adaptive hyper-parameter selection algorithm based on the multi-step grid search is employed in terms of the reduced features. Sequentially, the optimized ensemble submodels with their weighting strategies are adaptively determined to build the final SEN model. The proposed method is verified by using benchmark nearinfrared data, high dimensional mechanical frequency spectrum data and industrial dioxin emission concentration data.

show abstract

Least squares support vector machine with self-organizing multiple kernel learning and sparsity

Cited by 32 publications

References 42 publications

Predicting Students’ Academic Performance with Conditional Generative Adversarial Network and Deep SVM

Predicting Students’ Academic Performance with Conditional Generative Adversarial Network and Deep SVM

Long-term load forecasting using grey wolf optimizer -least-squares support vector machine

Multisource Latent Feature Selective Ensemble Modeling Approach for Small-Sample High-Dimensional Process Data in Applications

Contact Info

Product

Resources

About