Precise Multi-Class Classification of Brain Tumor via Optimization Based Relevance Vector Machine

The objective of this research is to examine the use of feature selection and classification methods for distinguishing different types of brain tumors.The brain tumor is characterized by an anomalous proliferation of brain cells that can either be benign or malignant.Most tumors are misdiagnosed due to the variabil-ity and complexity of lesions,which reduces the survival rate in patients.Diagno-sis of brain tumors via computer vision algorithms is a challenging task.Segmentation and classification of brain tumors are currently one of the most essential surgical and pharmaceutical procedures.Traditional brain tumor identi-fication techniques require manual segmentation or handcrafted feature extraction that is error-prone and time-consuming.Hence the proposed research work is mainly focused on medical image processing,which takes Magnetic Resonance Imaging(MRI)images as input and performs preprocessing,segmentation,fea-ture extraction,feature selection,similarity measurement,and classification steps for identifying brain tumors.Initially,the medianfilter is practically applied to the input image to reduce the noise.The graph-cut segmentation technique is used to segment the tumor region.The texture feature is extracted from the output of the segmented image.The extracted feature is selected by using the Ant Colony Opti-mization(ACO)algorithm to improve the performance of the classifier.This prob-abilistic approach is used to solve computing issues.The Euclidean distance is used to calculate the degree of similarity for each extracted feature.The selected feature value is given to the Relevance Vector Machine(RVM)which is a multi-class classification technique.Finally,the tumor is classified as abnormal or nor-mal.The experimental result reveals that the proposed RVM technique gives a better accuracy range of 98.87%when compared to the traditional Support Vector Machine(SVM)technique.

Estimation of illumination chromaticity via adaptive reduced relevance vector machine

A new regression algorithm of an adaptive reduced relevance vector machine is proposed to estimate the illumination chromaticity of an image for the purpose of color constancy. Within the framework of sparse Bayesian learning, the algorithm extends the relevance vector machine by combining global and local kernels adaptively in the form of multiple kernels, and the improved locality preserving projection （LLP） is then applied to reduce the column dimension of the multiple kernel input matrix to achieve less training time. To estimate the illumination chromaticity, the algorithm is trained by fuzzy central values of chromaticity histograms of a set of images and the corresponding illuminants. Experiments with real images indicate that the proposed algorithm performs better than the support vector machine and the relevance vector machine while requiring less training time than the relevance vector machine.

NEW METHOD FOR WHITE BLOOD CELL DETECTION BASED ON RELEVANCE VECTOR MACHINE

A new method for the white blood cell （WBC） detection is presented based on the relevance vector machine （RVM）. Firstly,the sparse relevance vectors （RVs） are obtained while fitting the 1-D histogram by RVM. Then,the needed threshold value is directly selected from these limited RVs. Finally,the entire connective WBC regions are segmented from the original image. The method is used for the WBC detection. It reduces the interference induced by the illumination and the staining. It has advantages of the high computation efficiency and the no extra parameter setting. Experimental results demonstrate good performances of the method.

Seismic liquefaction potential assessment by using relevance vector machine

Determining the liquefaction potential of soil is important in earthquake engineering. This study proposes the use of the Relevance Vector Machine （RVM） to determine the liquefaction potential of soil by using actual cone penetration test （CPT） data. RVM is based on a Bayesian formulation of a linear model with an appropriate prior that results in a sparse representation. The results are compared with a widely used artificial neural network （ANN） model. Overall, the RVM shows good performance and is proven to be more accurate than the ANN model. It also provides probabilistic output. The model provides a viable tool for earthquake engineers to assess seismic conditions for sites that are susceptible to liquefaction.

Relevance vector machine technique for the inverse scattering problem

A novel method based on the relevance vector machine（RVM） for the inverse scattering problem is presented in this paper.The nonlinearity and the ill-posedness inherent in this problem are simultaneously considered.The nonlinearity is embodied in the relation between the scattered field and the target property,which can be obtained through the RVM training process.Besides,rather than utilizing regularization,the ill-posed nature of the inversion is naturally accounted for because the RVM can produce a probabilistic output.Simulation results reveal that the proposed RVM-based approach can provide comparative performances in terms of accuracy,convergence,robustness,generalization,and improved performance in terms of sparse property in comparison with the support vector machine（SVM） based approach.

Data-driven optimal operation of the industrial methanol to olefin process based on relevance vector machine

Methanol to olefin(MTO)technology provides the opportunity to produce olefins from nonpetroleum sources such as coal,biomass and natural gas.More than 20 commercial MTO plants have been put into operation.Till now,contributions on optimal operation of industrial MTO plants from a process systems engineering perspective are rare.Based on relevance vector machine(RVM),a data-driven framework for optimal operation of the industrial MTO process is established to fully utilize the plentiful industrial data sets.RVM correlates the yield distribution prediction of main products and the operation conditions.These correlations then serve as the constraints for the multi-objective optimization model to pursue the optimal operation of the plant.Nondominated sorting genetic algorithmⅡis used to solve the optimization problem.Comprehensive tests demonstrate that the ethylene yield is effectively improved based on the proposed framework.Since RVM does provide the distribution prediction instead of point estimation,the established model is expected to provide guidance for actual production operations under uncertainty.

MULTIPLE KERNEL RELEVANCE VECTOR MACHINE FOR GEOSPATIAL OBJECTS DETECTION IN HIGH-RESOLUTION REMOTE SENSING IMAGES

Geospatial objects detection within complex environment is a challenging problem in remote sensing area. In this paper, we derive an extension of the Relevance Vector Machine (RVM) technique to multiple kernel version. The proposed method learns an optimal kernel combination and the associated classifier simultaneously. Two feature types are extracted from images, forming basis kernels. Then these basis kernels are weighted combined and resulted the composite kernel exploits interesting points and appearance information of objects simultaneously. Weights and the detection model are finally learnt by a new algorithm. Experimental results show that the proposed method improve detection accuracy to above 88%, yields good interpretation for the selected subset of features and appears sparser than traditional single-kernel RVMs.

Prediction of Compressive Strength of Various SCC Mixes Using Relevance Vector Machine

This paper discusses the applicability of relevance vector machine(RVM)based regression to predict the compressive strength of various self compacting concrete(SCC)mixes.Compressive strength data various SCC mixes has been consolidated by considering the effect of water cement ratio,water binder ratio and steel fibres.Relevance vector machine(RVM)is a machine learning technique that uses Bayesian inference to obtain parsimonious solutions for regression and classification.The RVM has an identical functional form to the support vector machine,but provides probabilistic classification and regression.RVM is based on a Bayesian formulation of a linear model with an appropriate prior that results in a sparse representation.Compressive strength model has been developed by using MATLAB software for training and prediction.About 75%of the data has been used for development of model and 30%of the data is used for validation.The predicted compressive strength for SCC mixes is found to be in very good agreement with those of the corresponding experimental observations available in the literature.

Fault Detection and Recovery for Full Range of Hydrogen Sensor Based on Relevance Vector Machine

In order to improve the reliability of hydrogen sensor,a novel strategy for full range of hydrogen sensor fault detection and recovery is proposed in this paper. Three kinds of sensors are integrated to realize the measurement for full range of hydrogen concentration based on relevance vector machine( RVM). Failure detection of hydrogen sensor is carried out by using the variance detection method. When a sensor fault is detected,the other fault-free sensors can recover the fault data in real-time by using RVM predictor accounting for the relevance of sensor data. Analysis,together with both simulated and experimental results,a full-range hydrogen detection and hydrogen sensor self-validating experiment is presented to demonstrate that the proposed strategy is superior at accuracy and runtime compared with the conventional methods. Results show that the proposed methodology provides a better solution to the full range of hydrogen detection and the reliability improvement of hydrogen sensor.

Feature Selection by Merging Sequential Bidirectional Search into Relevance Vector Machine in Condition Monitoring

For more accurate fault detection and diagnosis, there is an increasing trend to use a large number of sensors and to collect data at high frequency. This inevitably produces large-scale data and causes difficulties in fault classification. Actually, the classification methods are simply intractable when applied to high-dimensional condition monitoring data. In order to solve the problem, engineers have to resort to complicated feature extraction methods to reduce the dimensionality of data. However, the features transformed by the methods cannot be understood by the engineers due to a loss of the original engineering meaning. In this paper, other forms of dimensionality reduction technique（feature selection methods） are employed to identify machinery condition, based only on frequency spectrum data. Feature selection methods are usually divided into three main types： filter, wrapper and embedded methods. Most studies are mainly focused on the first two types, whilst the development and application of the embedded feature selection methods are very limited. This paper attempts to explore a novel embedded method. The method is formed by merging a sequential bidirectional search algorithm into scale parameters tuning within a kernel function in the relevance vector machine. To demonstrate the potential for applying the method to machinery fault diagnosis, the method is implemented to rolling bearing experimental data. The results obtained by using the method are consistent with the theoretical interpretation, proving that this algorithm has important engineering significance in revealing the correlation between the faults and relevant frequency features. The proposed method is a theoretical extension of relevance vector machine, and provides an effective solution to detect the fault-related frequency components with high efficiency.

Seismic fragility analysis of bridges by relevance vector machine based demand prediction model

A relevance vector machine(RVM)based demand prediction model is explored for efficient seismic fragility analysis(SFA)of a bridge structure.The proposed RVM model integrates both record-to-record variations of ground motions and uncertainties of parameters characterizing the bridge model.For efficient fragility computation,ground motion intensity is included as an added dimension to the demand prediction model.To incorporate different sources of uncertainty,random realizations of different structural parameters are generated using Latin hypercube sampling technique.Mean fragility,along with its dispersions,is estimated based on the log-normal fragility model for different critical components of a bridge.The effectiveness of the proposed RVM model-based SFA of a bridge structure is elucidated numerically by comparing it with fragility results obtained by the commonly used SFA approaches,while considering the most accurate direct Monte Carlo simulation-based fragility estimates as the benchmark.The proposed RVM model provides a more accurate estimate of fragility than conventional approaches,with significantly less computational effort.In addition,the proposed model provides a measure of uncertainty in fragility estimates by constructing confidence intervals for the fragility curves.

Spatio-Temporal Prediction of Root Zone Soil Moisture Using Multivariate Relevance Vector Machines

Root zone soil moisture at one and two meter depths are forecasted four days into the future. In this article, we propose a new multivariate output prediction approach to root zone soil moisture assessment using learning machine models. These models are known for their robustness, efficiency, and sparseness;they provide a statistically sound approach to solving the inverse problem and thus to building statistical models. The multivariate relevance vector machine (MVRVM) is used to build a model that forecasts soil moisture states based upon current soil moisture and soil temperature conditions. The methodology combines the data at different depths from 5 cm to 50 cm, the largest of which corresponds to the depth at which the soil moisture sensors are generally operational, to produce soil moisture predictions at larger depths. The MVRVM test results for soil moisture predictions at 1 m and 2 m depth on the 4th day are excellent with RMSE = 0.0131 m3/m3 for 1 m;and RMSE = 0.0015 m3/m3 for 2 m forecasted values. The statistics of predictions for 4th day (CoE = 0.87 for 1 m and CoE = 0.96 for 2 m) indicate good model generalization capability and computations show good agreement with actual measurements with R2 = 0.88 and R2 = 0.97 for 1 m and 2 m depths, respectively. The MVRVM produces good results for all four days. Bootstrapping is used to check over/under-fitting and uncertainty in model estimates.

基于SSA-RVM的滚动轴承可靠度评估与预测

为解决滚动轴承运行可靠度预测问题,这里提出了基于麻雀搜索算法-相关向量机的滚动轴承可靠度预测方法。首先对轴承振动信号在时域、频域及时频域构成的维数较高的向量集利用主成分分析算法进行降维;然后将降维后的特征集作为轴承的退化状态特征输入到逻辑回归模型中,进行滚动轴承可靠性评估;然后将轴承的性能退化状态特征作为麻雀搜索算法-相关向量机模型的输入,获取预测结果;最终把结果带入到逻辑回归模型中,预测轴承的运行可靠度。实验结果表明提出的算法在预测滚动轴承运行可靠性中具有明显优势。

Support Vector Machine active learning for 3D model retrieval

In this paper, we present a novel Support Vector Machine active learning algorithm for effective 3D model retrieval using the concept of relevance feedback. The proposed method learns from the most informative objects which are marked by the user, and then creates a boundary separating the relevant models from irrelevant ones. What it needs is only a small number of 3D models labelled by the user. It can grasp the user＇s semantic knowledge rapidly and accurately. Experimental results showed that the proposed algorithm significantly improves the retrieval effectiveness. Compared with four state-of-the-art query refinement schemes for 3D model retrieval, it provides superior retrieval performance after no more than two rounds of relevance feedback.

A new support vector machine based multiuser detection scheme

In order to suppress the multiple access interference (MAI) in 3G, which limits the capacity of a CDMA communication system, a fast relevance vector machine (FRVM) is employed in the multiuser detection (MUD) scheme. This method aims to overcome the shortcomings of many ordinary support vector machine (SVM) based MUD schemes, such as the long training time and the inaccuracy of the decision data, and enhance the performance of a CDMA communication system. Computer simulation results demonstrate that the proposed FRVM based multiuser detection has lower bit error rate, costs short training time, needs fewer kernel functions and possesses better near-far resistance.

基于POA-RVM模型的抽蓄机组故障诊断

有效的故障诊断方法不仅能快速、准确地辨别抽蓄机组故障类型,还能降低抽水蓄能电站的运行维护成本。针对相关向量机(RVM)有关参数的调整不当导致诊断结果受影响的问题,提出利用鹈鹕优化算法(POA)对相关向量机参数的选取进行优化,构建鹈鹕优化算法和相关向量机组合的分类模型(POA-RVM)。选取仙居抽水蓄能电站4台抽蓄机组在5种状态下的数据进行预处理和特征选取后构成故障样本集,并分别采用标准相关向量机,以及用遗传算法、粒子群算法和灰狼算法优化的相关向量机模型对这些故障样本进行分类。结果表明:与标准相关向量机,以及经遗传算法、粒子群算法和灰狼算法优化的相关向量机模型相比,POA-RVM模型有效提高了抽蓄机组故障诊断的准确率。

基于改进LFQPSO优化MRVM的轴向柱塞泵故障诊断

针对传统粒子群优化算法以准确率或误判率作为适应度函数耗时长和轴向柱塞泵故障机制较为复杂的问题,提出一种基于改进适应度函数的Lévy飞行量子粒子群优化(QPSO)多分类相关向量机(MRVM)的轴向柱塞泵概率性智能软状态判别方法。为了克服人为设定核参数不精确、效率低等缺点,采用基于Lévy飞行的QPSO搜索MRVM的最优核参数;为了缩短寻优时间,将样本间余弦相似度作为寻优算法的适应度函数,并利用UCI机器学习标准数据集进行仿真来验证改进后优化方法的有效性及优越性;采集柱塞泵不同故障状态的数据,提取时频域和时域特征,输入到优化后的MRVM中,进行训练及测试。实验结果表明:所提方法可以有效提高故障诊断的准确率及诊断效率,同时能够实现软分类,即以概率形式输出诊断结果,能够为设备检修及维护提供可靠且符合实际的故障信息。

基于PCA-RVM的矿山岩石爆破粒径预测模型

为解决露天矿山爆破开采过程中岩石爆破粒径大小难以获取的问题,提出一种基于主成分分析法(PCA)及相关向量机(RVM)相结合的矿山岩石爆破粒径预测模型.该模型利用PCA对样本数据进行降维处理,选取出4个相互独立的主成分变量,并借助RVM构建主成分与爆破粒径之间的非线性映射关系,从而建立预测模型.将该模型应用于工程实例,并与BP神经网络和LM双隐含层模型进行对比.结果表明,在相同学习样本下,PCA-RVM模型预测结果与实际值更加接近,在平均相对误差和均方差上远小于另两种模型.

基于PSO-DSRVM的边坡变形预测

为了建立高精度的边坡位移预测模型,文章采用基于粒子群优化(PSO)的双稀疏相关向量机(DSRVM)建立边坡稳定性和影响因素之间的非线性关系。双稀疏相关向量机是在变分和相关向量机(RVM)框架下提出的一种多核组合优化的方法,相比于RVM和其他多核学习方法,DSRVM不仅有更少的训练时间,并且能够得到更高的预测精度。由于DSRVM的核参数对预测效果的影响较大,文章采用粒子群算法实现多个核参数的优化选取并应用于边坡位移预测。最后将本文提出的基于粒子群优化的双稀疏相关向量机(PSO-DSRVM)预测结果与极限学习机(ELM)和小波神经网络(WNN)预测结果进行对比,通过均方根误差(RMSE)、复相关系数(R^(2))和平均相对预测误差(ARPE)进行评价,验证了PSO-DSRVM模型在边坡变形预测上的可行性。

基于特征融合与HPO-RVM的滚动轴承剩余寿命预测

为准确预测轴承的剩余使用寿命,提出基于特征融合与猎食者-猎物优化(HPO)算法优化相关向量机的轴承剩余寿命预测方法。提取时域、频域和时频域特征准确描述轴承的退化状态,利用综合评价指标对提取的特征进行筛选得到敏感特征集;采用核熵成分分析对敏感特征进行自适应融合,得到轴承的退化特征;构建混合核函数作为相关向量机的核函数以提高模型预测性能;最后,利用HPO算法得到混合核函数的参数,将寻优得到的参数用于寿命预测模型的训练。通过对轴承加速退化数据集进行实验,结果表明:所构建的寿命预测模型优于BP、ELM、SVM等模型,构造的混合核函数模型优于高斯核函数模型,采用的优化算法优于粒子群、遗传算法等。