Feature weighting and SVM parameters optimization based on genetic algorithms for classification problems

Phan, Anh Viet; Nguyen, Minh Le; Bui, Lam Thu

doi:10.1007/s10489-016-0843-6

Cited by 97 publications

(45 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For compactness, each confusion matrix is condensed into a single number, called the accuracy [33]. The accuracy can be calculated by averaging the true positive rates across six classes and, in this paper, it was employed as the measure of performance.…”

Section: Resultsmentioning

confidence: 99%

A New Terrain Classification Framework Using Proprioceptive Sensors for Mobile Robots

Zhao

Dong

2017

Mathematical Problems in Engineering

View full text Add to dashboard Cite

Mobile robots that operate in real-world environments interact with the surroundings to generate complex acoustics and vibration signals, which carry rich information about the terrain. This paper presents a new terrain classification framework that utilizes both acoustics and vibration signals resulting from the robot-terrain interaction. As an alternative to handcrafted domain-specific feature extraction, a two-stage feature selection method combining ReliefF and mRMR algorithms was developed to select optimal feature subsets that carry more discriminative information. As different data sources can provide complementary information, a multiclassifier combination method was proposed by considering a priori knowledge and fusing predictions from five data sources: one acoustic data source and four vibration data sources. In this study, four conceptually different classifiers were employed to perform the classification, each with a different number of optimal features. Signals were collected using a tracked robot moving at three different speeds on six different terrains. The new framework successfully improved classification performance of different classifiers using the newly developed optimal feature subsets. The greater improvement was observed for robot traversing at lower speeds.

show abstract

Section: Resultsmentioning

confidence: 99%

A New Terrain Classification Framework Using Proprioceptive Sensors for Mobile Robots

Zhao

Dong

2017

Mathematical Problems in Engineering

View full text Add to dashboard Cite

show abstract

“…When c is small, there are more samples between the two boundaries. The possibility of misclassification becomes larger, and the fit to the sample is reduced, but it may be more reasonable because there may be noise between the samples (Bao, Hu, & Xiong, 2013;Phan, Nguyen, & Bui, 2017). c is a key factor in determining the SVM learning ability and experience risk coordination.…”

Section: Support Vector Machine (Svm)mentioning

confidence: 99%

An intelligent fault identification method of rolling bearings based on SVM optimized by improved GWO

Huang

Fan

Wei

et al. 2019

Systems Science & Control Engineering

View full text Add to dashboard Cite

To improve the accuracy and the recognition efficiency of a bearing fault diagnosis, a fault diagnosis method based upon the improved grey wolf optimization (IGWO) algorithm and support vector machine (SVM) is proposed in the following manners. First, the data are pre-processed by using the set ensemble empirical mode decomposition (EEMD), Shannon wavelet packet entropy (SWPE), and principal component analysis (PCA). Next, the idea of updating the host bird nest by the cuckoo search (CS) optimization algorithm is introduced into the grey wolf optimization (GWO) algorithm to obtain the IGWO algorithm. Then, the SVM is optimized by the IGWO algorithm to obtain optimal parameters for a new diagnostic model. This model improves the problem where the algorithm easily to falls into a local optimum. The learning ability and the generalization ability of the SVM are also enhanced. Finally, the effectiveness of the optimization model is tested by two different bearing data sets. The results show that compared to the genetic algorithm (GA), particle swarm optimization (PSO) and GWO algorithm optimization, the IGWO algorithm can be more accurate and efficient when diagnosing bearings. ARTICLE HISTORY

show abstract

“…The superiority of SVM detection performance is related to the performance of the entire detection system. The classification effect of the SVM classifier depends largely on the reasonable selection of the penalty factor C and the parameters in the kernel function [3].…”

Section: A Svm Parameter Impact Analysismentioning

confidence: 99%

Electric Larceny Detection Based on Support Vector Machine

Li¹,

Zeng²,

Ye³

et al. 2019

Proceedings of the 2019 International Conference on Precision Machining, Non-Traditional Machining and Intelligent Manufacturin

View full text Add to dashboard Cite

The design and application of power system line loss calculation and lean management system have important guiding significance in guiding loss reduction and energy saving and promoting line loss management. In recent years, the electric energy data acquire system, as a tool that can effectively meet the power enterprise's demand for power consumption information, has also accumulated a large amount of user power consumption data while meeting the power supply marketing automation needs. These power consumption data contain huge user power usage information. Therefore, the user data collected by the power electric energy data acquire system can be analyzed and processed to identify users with high suspicion of power severance, so as to reduce the management line loss. To this end, this paper studies a small-volume user anomaly power detection scheme based on Support Vector Machine (SVM), which can effectively identify the abnormal power consumption mode by tracking and screening the load data of the user for a period of time. An unbalanced sample synthesis processing model based on SMOTE+Bagging is constructed. The differential evolution algorithm is used to optimize the SVM parameters, which solves the problem that SVM classification performance is more affected by parameters. At the same time, the operational efficiency of the SVM-based Bagging integrated classification model is guaranteed.

show abstract

Feature weighting and SVM parameters optimization based on genetic algorithms for classification problems

Cited by 97 publications

References 28 publications

A New Terrain Classification Framework Using Proprioceptive Sensors for Mobile Robots

A New Terrain Classification Framework Using Proprioceptive Sensors for Mobile Robots

An intelligent fault identification method of rolling bearings based on SVM optimized by improved GWO

Electric Larceny Detection Based on Support Vector Machine

Contact Info

Product

Resources

About