Integrating Local and Global Error Statistics for Multi-Scale RBF Network Training: An Assessment on Remote Sensing Data

Mountrakis, Giorgos; Zhuang, Wei

doi:10.1371/journal.pone.0040093

Cited by 3 publications

(3 citation statements)

References 67 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The core of classifier construction is to select the correct kernel function. The Radial Basis Function (RBF) kernel can project nonlinear matters into high-latitude space and then establish linear functions to solve nonlinear issues [ 16 ]. The calculation method of the RBF kernel is shown in equation ( 6 ), where γ represents the distance of the horizontal axis:

…”

Section: Methodsmentioning

confidence: 99%

Effect of Comprehensive Rehabilitation Training Program in Orthopedic Nursing of Patients with Residual Limb Injury Caused by Crush

Zhou

2022

Journal of Healthcare Engineering

View full text Add to dashboard Cite

This study was developed to explore the role and application value of a comprehensive rehabilitation training (CRT) program based on the remote monitoring system of limb rehabilitation training (LRT-RM system) in the rehabilitation nursing of patients with residual limb injuries caused by crush. The LRT-RM system was constructed based on the characteristics of limb movement and using the time-domain analysis method and support vector machine (SVM). The 84 crush injury patients were selected as the research objects and divided into a control group (Con group, received conventional rehabilitation therapy) and a CRT group (received conventional rehabilitation therapy + functional training) according to different therapies, with 42 people in each group. The incidence of compound injuries and the incidence of residual limb injuries were counted and compared for patients in two groups. The differences in renal function, blood electrolytes, and biochemical indicators before and after treatment were analyzed. The MOS 36-item short-form health survey (SF-36) scale was selected to evaluate the improvement of physical and mental health of the patients before treatment and 1 month (time point (TP1)), 3 months (TP2), 6 months (TP3), and 12 months (TP4) after the treatment. It was found that, after the intervention, the values of serum creatinine (Scr), blood urea nitrogen (BUN), uric acid (UA), K+, P3+, and white blood cells (WBC) of patients in CRT group were obviously lower than those of Con group ( P < 0.05), and the values of carbon dioxide combining power (CO2CP), Ca2+, hemoglobin (Hb), red blood cell (RBC), total protein (TP), and albumin (ALB) were obviously higher than the values in Con group ( P < 0.05). In the CRT group, the residual limb injury rate was lower in elbow, wrist, shoulder joint, ankle joint, and toe ( P < 0.05) and extremely lower in knee joint in contrast to that in the Con group ( P < 0.001). The score of SF-36 was dramatically higher than that in the Con group ( P < 0.05). It suggested that the CRT program based on the LRT-RM system was helpful for the rehabilitation of patients with crush injuries, and it can reduce the incidence of residual limb injuries in patients. Results of this study could provide a reference basis for the treatment of residual limb injuries caused by crush.

show abstract

…”

Section: Methodsmentioning

confidence: 99%

Effect of Comprehensive Rehabilitation Training Program in Orthopedic Nursing of Patients with Residual Limb Injury Caused by Crush

Zhou

2022

Journal of Healthcare Engineering

View full text Add to dashboard Cite

show abstract

“…Second, we also compared the predictive results of the proposed method with that of four other commonly used classifiers, i.e., Naïve Bayes 20 , Logistic Function 21 , RBFNetwork 22 , and Random Forest 23 as implemented in WEKA 24 . The jackknife test results for identifying m 6 A sites in the benchmark dataset for different classifiers were listed in Table 3 .…”

Section: Resultsmentioning

confidence: 99%

Identification and analysis of the N6-methyladenosine in the Saccharomyces cerevisiae transcriptome

Chen

Tran

Zhou

et al. 2015

Sci Rep

102

View full text Add to dashboard Cite

Knowledge of the distribution of N6-methyladenosine (m6A) is invaluable for understanding RNA biological functions. However, limitation in experimental methods impedes the progress towards the identification of m6A site. As a complement of experimental methods, a support vector machine based-method is proposed to identify m6A sites in Saccharomyces cerevisiae genome. In this model, RNA sequences are encoded by their nucleotide chemical property and accumulated nucleotide frequency information. It is observed in the jackknife test that the accuracy achieved by the proposed model in identifying the m6A site was 78.15%. For the convenience of experimental scientists, a web-server for the proposed model is provided at http://lin.uestc.edu.cn/server/m6Apred.php.

show abstract

“…When used for classifying problems, there are three important factors for evaluating network performance: 1) classifying accuracy, 2) network size, and 3) training time. To achieve good network performance, different optimization algorithms are used to train the RBF hidden layer, such as K-means clustering [1, 2], fuzzy C-means clustering [3, 4], fuzzy K-nearest neighbors [5], differential evolution [6, 7], and other optimization algorithms [8–12]. However, in most of these methods, the number of RBF hidden nodes is assigned a priori, which may lead to poor adaptability for different sample sets.…”

Section: Introductionmentioning

confidence: 99%

A Structure-Adaptive Hybrid RBF-BP Classifier with an Optimized Learning Strategy

2016

View full text Add to dashboard Cite

This paper presents a structure-adaptive hybrid RBF-BP (SAHRBF-BP) classifier with an optimized learning strategy. SAHRBF-BP is composed of a structure-adaptive RBF network and a BP network of cascade, where the number of RBF hidden nodes is adjusted adaptively according to the distribution of sample space, the adaptive RBF network is used for nonlinear kernel mapping and the BP network is used for nonlinear classification. The optimized learning strategy is as follows: firstly, a potential function is introduced into training sample space to adaptively determine the number of initial RBF hidden nodes and node parameters, and a form of heterogeneous samples repulsive force is designed to further optimize each generated RBF hidden node parameters, the optimized structure-adaptive RBF network is used for adaptively nonlinear mapping the sample space; then, according to the number of adaptively generated RBF hidden nodes, the number of subsequent BP input nodes can be determined, and the overall SAHRBF-BP classifier is built up; finally, different training sample sets are used to train the BP network parameters in SAHRBF-BP. Compared with other algorithms applied to different data sets, experiments show the superiority of SAHRBF-BP. Especially on most low dimensional and large number of data sets, the classification performance of SAHRBF-BP outperforms other training SLFNs algorithms.

show abstract

Integrating Local and Global Error Statistics for Multi-Scale RBF Network Training: An Assessment on Remote Sensing Data

Cited by 3 publications

References 67 publications

Effect of Comprehensive Rehabilitation Training Program in Orthopedic Nursing of Patients with Residual Limb Injury Caused by Crush

Effect of Comprehensive Rehabilitation Training Program in Orthopedic Nursing of Patients with Residual Limb Injury Caused by Crush

Identification and analysis of the N6-methyladenosine in the Saccharomyces cerevisiae transcriptome

A Structure-Adaptive Hybrid RBF-BP Classifier with an Optimized Learning Strategy

Contact Info

Product

Resources

About