Cross-Project Software Defect Prediction Based on SMOTE and Deep Canonical Correlation Analysis

Cross-Project Defect Prediction(CPDP)is a method that utilizes historical data from other source projects to train predictive models for defect prediction in the target project.However,existing CPDP methods only consider linear correlations between features(indicators)of the source and target projects.These models are not capable of evaluating non-linear correlations between features when they exist,for example,when there are differences in data distributions between the source and target projects.As a result,the performance of such CPDP models is compromised.In this paper,this paper proposes a novel CPDP method based on Synthetic Minority Oversampling Technique(SMOTE)and Deep Canonical Correlation Analysis(DCCA),referred to as S-DCCA.Canonical Correlation Analysis(CCA)is employed to address the issue of non-linear correlations between features of the source and target projects.S-DCCA extends CCA by incorporating the MlpNet model for feature extraction from the dataset.The redundant features are then eliminated by maximizing the correlated feature subset using the CCA loss function.Finally,cross-project defect prediction is achieved through the application of the SMOTE data sampling technique.Area Under Curve(AUC)and F1 scores(F1)are used as evaluation metrics.This paper conducted experiments on 27 projects from four public datasets to validate the proposed method.The results demonstrate that,on average,our method outperforms all baseline approaches by at least 1.2%in AUC and 5.5%in F1 score.This indicates that the proposed method exhibits favorable performance characteristics.

Canonical Correlation Analysis of Agronomic Characters of Brassica juncea in Western China

[Objective] The study aimed at exploring the relationship among the agronomic characters of B. juncea in western China, in order to provide scientific basis for the breeding of B. juncea in western China. [Method] 39 B. juncea materials from western China were used for the canonical correlation analysis, and canonical correlations between each pair of the four ecological character （containing 18 variables） were verified, including yield characters （5 variables）, caulis characters （6 variables）, branch characters （3 variables） and pod characters （3 variables）. [Result] Yield per plant of B. juncea in western China suffered a tremendous influence from effective pod number per plant while was not significantly affected by the total pod number per plant, seed number per pod and 1 000-seed weight; the most important character related with the yield character of B. juncea in western China was caulis character, followed by the branch character and pod character; yield characters, caulis characters, branch characters and pod characters of B. juncea in western China were closely correlated. [Conclusion] In order to improve the yield characters of B. juncea in western China, caulis characters should be focused on, followed by branch characters and pod characters; rapeseed varieties with high performance in total pod number per plant and effective pod number per plant should be chosen through the perspectives of effective branch number, plant height, pod number of main inflorescence, fruit stalk number of main inflorescence and other traits, while rapeseed varieties with high performance in seed number per pod and 1 000-seed weight should be chosen through the perspectives of beak length and other traits.

Feature Fusion Multi-View Hashing Based on Random Kernel Canonical Correlation Analysis

Hashing technology has the advantages of reducing data storage and improving the efficiency of the learning system,making it more and more widely used in image retrieval.Multi-view data describes image information more comprehensively than traditional methods using a single-view.How to use hashing to combine multi-view data for image retrieval is still a challenge.In this paper,a multi-view fusion hashing method based on RKCCA(Random Kernel Canonical Correlation Analysis)is proposed.In order to describe image content more accurately,we use deep learning dense convolutional network feature DenseNet to construct multi-view by combining GIST feature or BoW_SIFT(Bag-of-Words model+SIFT feature)feature.This algorithm uses RKCCA method to fuse multi-view features to construct association features and apply them to image retrieval.The algorithm generates binary hash code with minimal distortion error by designing quantization regularization terms.A large number of experiments on benchmark datasets show that this method is superior to other multi-view hashing methods.

Study on soil water characteristics of tobacco fields based on canonical correlation analysis

In order to identify the principal factors influencing soil water characteristics （SWC） and evaluate SWC effectively, the multivariate-statistical canonical correlation analysis （CCA） method was used to study and analyze the correlation between SWC and soil physical and chemical properties. Twenty-two soil samples were taken from 11 main tobacco-growing areas in Guizhou Province in China and the soil water characteristic curves （SWCC） and basic physical and chemical properties of the soil samples were determined. The results show that： （1） The soil bulk density, soil total porosity and soil capillary porosity have significant effects on SWC of tobacco fiels. Bulk density and total porosity are positively correlated with soil water retention characteristics （SWRC）, and soil capillary porosity is positively correlated with soil water supply characteristics （SWSC）. （2） Soil samples from different soil layers at the same soil sampling point show similarity or consistency in SWC. Inadequate soil water supply capability and imbalance between SWRC and SWSC are problems of tobacco soil. （3） The SWC of loamy clay are generally superior to those of silty clay loam.

Novel DDoS Feature Representation Model Combining Deep Belief Network and Canonical Correlation Analysis

Distributed denial of service(DDoS)attacks launch more and more frequently and are more destructive.Feature representation as an important part of DDoS defense technology directly affects the efficiency of defense.Most DDoS feature extraction methods cannot fully utilize the information of the original data,resulting in the extracted features losing useful features.In this paper,a DDoS feature representation method based on deep belief network(DBN)is proposed.We quantify the original data by the size of the network flows,the distribution of IP addresses and ports,and the diversity of packet sizes of different protocols and train the DBN in an unsupervised manner by these quantified values.Two feedforward neural networks(FFNN)are initialized by the trained deep belief network,and one of the feedforward neural networks continues to be trained in a supervised manner.The canonical correlation analysis(CCA)method is used to fuse the features extracted by two feedforward neural networks per layer.Experiments show that compared with other methods,the proposed method can extract better features.

A NOVEL ALGORITHM FOR VOICE CONVERSION USING CANONICAL CORRELATION ANALYSIS

A novel algorithm for voice conversion is proposed in this paper. The mapping function of spectral vectors of the source and target speakers is calculated by the Canonical Correlation Analysis (CCA) estimation based on Gaussian mixture models. Since the spectral envelope feature remains a majority of second order statistical information contained in speech after Linear Prediction Coding (LPC) analysis, the CCA method is more suitable for spectral conversion than Minimum Mean Square Error (MMSE) because CCA explicitly considers the variance of each component of the spectral vectors during conversion procedure. Both objective evaluations and subjective listening tests are conducted. The experimental results demonstrate that the proposed scheme can achieve better per- formance than the previous method which uses MMSE estimation criterion.

ASSESSMENT OF INFLUENCE IN CANONICAL CORRELATION ANALYSIS

By expanding the perturbation of covariance matrix in the powers of er-ror term,the influence functions for five canonical measurements in CCA are devel-oped and three sample versions are given.For generalized correlation coefficient p_z,the influence function is a quadratic form of r.v.z,and its distribution is considered.A practical example iUustrates the utility of the proposed influence functions.

SPATIAL REGULARIZATION OF CANONICAL CORRELATION ANALYSIS FOR LOW-RESOLUTION FACE RECOGNITION

Canonical correlation analysis ( CCA ) based methods for low-resolution ( LR ) face recognition involve face images with different resolutions ( or multi-resolutions ), i.e.LR and high-resolution ( HR ) .For single-resolution face recognition , researchers have shown that utilizing spatial information is beneficial to improving the recognition accuracy , mainly because the pixels of each face are not independent but spatially correlated.However , for a multi-resolution scenario , there are no related works.Therefore , a method named spatial regularization of canonical correlation analysis ( SRCCA ) is developed for LR face recognition to improve the performance of CCA by the regularization utilizing spatial information of different resolution faces.Furthermore , the impact of LR and HR spatial regularization terms on LR face recognition is analyzed through experiments.

Canonical correlation analysis to land-use structure and its driving forces——Taking Yulin Prefecture as an example

In this paper, one of the most classical statistical methods, Canonical Correlation Analysis (CCA) is applied to identify quantitatively the driving forces of landuse structure in Yulin Prefecture. The main analysis is carried out through the software SPSS with the data on the level of towns and townships in 1992. The results indicate that landuse structure is determined by comprehensive action of different factors. Landuse structure with rural characteristics is mainly determined by geographical factors such as the elevation, temperature and precipitation, while the landuse structure with urban characteristics is mainly determined by demographic and socioeconomic conditions. At the same time, tests were carried out through the canonical correlation coefficient and redundancy analysis.

ON USING NON-LINEAR CANONICAL CORRELATION ANALYSIS FOR VOICE CONVERSION BASED ON GAUSSIAN MIXTURE MODEL

Voice conversion algorithm aims to provide high level of similarity to the target voice with an acceptable level of quality.The main object of this paper was to build a nonlinear relationship between the parameters for the acoustical features of source and target speaker using Non-Linear Canonical Correlation Analysis(NLCCA) based on jointed Gaussian mixture model.Speaker indi-viduality transformation was achieved mainly by altering vocal tract characteristics represented by Line Spectral Frequencies(LSF).To obtain the transformed speech which sounded more like the target voices,prosody modification is involved through residual prediction.Both objective and subjective evaluations were conducted.The experimental results demonstrated that our proposed algorithm was effective and outperformed the conventional conversion method utilized by the Minimum Mean Square Error(MMSE) estimation.

Canonical Correlation Analysis and climate research

Correlation analysis as used by meteorologists and oceanographers is a tool for the analysisof the spacial or temporal variability of physical fields. In his notes, Dr. Hasselmann pro-posed to combine correlation analysis and linear regression analysis in climate prediction re-search. The main idea is to decompose the physical field into its principal oscillation patterns.

Multiple moving sources passive location based on multiset canonical correlation analysis

To solve the problem of multiple moving sources passive location, a novel blind source separa- tion （BSS） algorithm based on the muhiset canonical correlation analysis （MCCA） is presented by exploiting the different temporal structure of uncorrelated source signals first, and then on the basis of this algorithm, a novel multiple moving sources passive location method is proposed using time difference of arrival （TDOA） and frequency difference of arrival （FDOA） measurements. The key technique of this location method is TDOA and FDOA joint estimation, which is based on BSS. By blindly separating mixed signals from multiple moving sources, the multiple sources location problem can be translated to each source location in turn, and the effect of interference and noise can also he removed. The simulation results illustrate that the performance of the MCCA algorithm is very good with relatively light computation burden, and the location algorithm is relatively simple and effective.

基于PSD特征的FBCCA脑电信号识别方法

当前基于稳态视觉诱发电位(steady-state visual evoked potential,SSVEP)的脑机接口(brain-computer interfaces,BCIs)使用的都是单一识别算法,针对不同时间长度的识别准确率较低。提出了一种基于滤波器组的典型相关分析(filter bank canonical correlation analysis,FBCCA)与功率谱密度(power spectral density,PSD)分析相结合的SSVEP识别算法,可以提高SSVEP识别的普适性与准确率。该方法使用FBCCA寻找高相似度的参考频率信号,再通过多组PSD分析来锁定最终的响应频率,完成频率识别。该方法无需经过训练就能得到较高的识别准确率。实验结果表明:在刺激时长为1 s时,该方法能达到86.61%的准确率,比PSD分析方法提升了5.44%,比典型相关性分析方法(canonical correlation analysis,CCA)提升了10.38%的准确率,比FBCCA提升了8.86%的准确率。

基于MLCCA的高速列车牵引系统故障诊断

随着高速列车牵引系统的复杂性和智能性的提高,牵引系统的故障诊断越来越重要,故提出了一种基于多层典型相关分析(multi-layer canonical correlation analysis,MLCCA)的故障诊断方法。该方法对数据进行非线性预测,达到提高牵引系统的故障检测性能的目的,并引入最小绝对值收缩和选择算子(least absolute shrinkage and selection operator,LASSO)中的惩罚项来提高预测精度。根据高速列车传感器数据的性质,对数据进行分组分析,设置2组不同的统计量比较算法的检测效果。采集牵引系统传感器故障时的数据进行仿真实验,并将所提方法与常用方法进行对比,实验结果验证了所提方法的优越性。

基于CCA-ELM模型的国产LNG出厂价格中短期预测研究——以陕西省为例

考虑供需基本面因素和非基本面因素,构建CCA-ELM模型用于国产LNG出厂价格的预测。供需基本面因素包括LNG的产量、销量、库存、气温以及原料气成本,非基本面影响因素包括原油、汽油、柴油、煤炭等替代能源价格与东北亚天然气现货价格。通过典型相关性分析,研究各个影响因素对价格的作用程度。以10个影响因素的周度数据为研究对象,以LNG出厂价格的历史序列与其影响因素构建CCA-ELM神经网络预测模型。10个影响因素整体与LNG出厂价格的相关性较强,中国LNG出厂价格受能源市场的影响程度较高,受供需基本面的影响程度较低。兼顾LNG出厂价格历史数据与影响因素的CCA-ELM模型有效改进了时间序列神经网络的预测方法,提高了预测精度。

A Novel CCA-NMF Whitening Method for Practical Machine Learning Based Underwater Direction of Arrival Estimation

Underwater direction of arrival(DOA)estimation has always been a very challenging theoretical and practical problem.Due to the serious non-stationary,non-linear,and non-Gaussian characteristics,machine learning based DOA estimation methods trained on simulated Gaussian noised array data cannot be directly applied to actual underwater DOA estimation tasks.In order to deal with this problem,environmental data with no target echoes can be employed to analyze the non-Gaussian components.Then,the obtained information about non-Gaussian components can be used to whiten the array data.Based on these considerations,a novel practical sonar array whitening method was proposed.Specifically,based on a weak assumption that the non-Gaussian components in adjacent patches with and without target echoes are almost the same,canonical cor-relation analysis(CCA)and non-negative matrix factorization(NMF)techniques are employed for whitening the array data.With the whitened array data,machine learning based DOA estimation models trained on simulated Gaussian noised datasets can be used to perform underwater DOA estimation tasks.Experimental results illustrated that,using actual underwater datasets for testing with known machine learning based DOA estimation models,accurate and robust DOA estimation performance can be achieved by using the proposed whitening method in different underwater con-ditions.

基于VAE-OCCA的质量相关故障检测方法研究

由于闭环反馈系统的存在,并不是所有故障均会导致质量发生恶化。质量变量通常难以获得或具有一定的延迟,传统的无监督方法不能在检测过程是否正常的同时判断故障对质量的影响。典型相关分析(canonical correlation analysis,CCA)是一种经典的有监督方法,可以考虑输入输出间的关系,已被用于质量相关故障检测。然而,过程数据存在着维度高、非线性等问题,流程系统的复杂性使得CCA对于隐藏特征的捕获更具挑战性。提出了一种变分自编码器-正交典型相关分析(variational automatic encoder-orthogonal CCA,VAE-OCCA)方法。首先,利用变分自编码器对输入数据进行无监督自适应学习,实现对高维非线性过程变量的特征提取;进而,基于典型相关分析方法考虑输入输出关系,利用得到的相关系数矩阵进行奇异值分解建立质量相关和质量无关监测统计量;最后,通过工业案例测试说明提出方法的有效性及优越性。

Asymptotic distributions in the projection pursuit based canonical correlation analysis

In this paper, associations between two sets of random variables based on the projection pursuit (PP) method are studied. The asymptotic normal distributions of estimators of the PP based canonical correlations and weighting vectors are derived.

Canonical correlation analysis of hydrological response and soil erosion under moving rainfall

The impacts of rainfall direction on the degree of hydrological response to rainfall properties were investigated using comparative rainfall-runoff experiments on a small-scale slope(4 m×1 m),as well as canonical correlation analysis(CCA).The results of the CCA,based on the observed data showed that,under conditions of both upstream and downstream rainfall movements,the hydrological process can be divided into instantaneous and cumulative responses,for which the driving forces are rainfall intensity and total rainfall,and coupling with splash erosion and wash erosion,respectively.The response of peak runoff(Pr)to intensity-dominated rainfall action appeared to be the most significant,and also runoff(R)to rainfall-dominated action,both for upstream-and downstream-moving conditions.Furthermore,the responses of sediment erosion in downstream-moving condition were more significant than those in upstream-moving condition.This study indicated that a CCA between rainfall and hydrological characteristics is effective for further exploring the rainfall-runoff-erosion mechanism under conditions of moving rainfall,especially for the downstream movement condition.

Convergence rate of kernel canonical correlation analysis

Kernel canonical correlation analysis(CCA) is a nonlinear extension of CCA,which aims at extract-ing the information shared by two random variables. It has wide applications in many fields,such as information retrieval. This paper gives the convergence rate analysis of kernel CCA under some approximation conditions and some suggestions on how to choose the regularization parameter. The result shows that the convergence rate only depends on two parameters:the rate of regularization parameter and the decay rate of eigenvalues of compact operator VY X,and it gives better understanding of kernel CCA.