Papers by Keyword: PCA

Abstract: The development of information technology gives rise to explosive growth of the amount of data. As a result, a more effective data mining method in pattern recognition is called into existence, which can properly reflect the inherent daily activity structure of metro travelers. This study is aimed to enrich the traditional clustering methods and provide practical information in dealing with traffic volume variation to the metro system operations. In this study, daily metro origin-destination (OD) data come from smart card records of Shenzhen, China, which cover 290 days and 118 stations. Principal component analysis (PCA) and singular value decomposition (SVD) are applied to conduct dimensionality reduction. Affinity propagation is then chosen to cluster the dimensionality reduced matrix to identify demand patterns of the metro OD matrix. Eleven representative categories are clustered and shown.

Abstract: Hydrometallurgy is a popular metallurgical technology. Filter press is common but vital to the production of hydrometallurgy. Hence, the process monitoring of filter press is of great significance for hydrometallurgy. Due to data analysis and related knowledge of filter press, Principal component analysis (PCA) is applied to process monitoring of the filter press via two traditional statistics. However, modeling and test data collected from actual production suffers from outliers, missing data, inconsistent sampling period between variables. Based on these practical problems, corresponding data proceeding technique is proposed. The final application simulation illustrates the validity of the proposed method.

Abstract: In this paper, we present an algorithm which detects human hand by skin color information in YCbCr and HIS color model. And for confirming special human hand we use circle rate of region to detect hand region because human hand have complex edge than other region, thus circle rate of hand region is usually more greater. For the recognition of detected hand, we use the Hausdorff to tracking the hand region. And we employed a recognition method based on PCA algorithm to recognize the hand gestures. The experimental results show that an algorithm plays an efficient effort for hand gesture recognition.

Abstract: This work forms part of large project for measuring the skin colors. This topic has been historically extensively studied due to the strong need from the photographic, digital imaging and medical applications. However there are still many unresolved issues, for example the measuring accuracy and the difference between different measuring methods. The paper focused on one of the measuring methods: camera. The goal is to develop PCA methods to reconstruct the reflectance from images captured by a camera, and the result shows that using three components is enough to acquire high accuracy, and it is possible to have a single skin model to predict all the available skin colors.

Abstract: T-2DPCA, a novel approach considering the third-order tensors as linear operators on the space of oriented matrices, benefits from treating a 2D image as an inherently integrated object, has been proposed recently and showed better performance than traditional matrix PCA in image analysis and recognition. In T-2DPCA, a reconstructing tubal coefficient is obtained from the defined tensor product, called T-product, of a 2D image and a 2D basis element. In this study, by assuming that an eigenvector of the covariance tensor of the 2D training images is the tensor linear combination, called T-linear combination, of the training images, the T-2DPCA is improved to a new version with better performance. The improved method is further extended to a nonlinear version by using the general kernel trick in machine learning field, but with a new inner product called inside product defined with the T-product in the third-order tensor spaces, and simultaneously the general inner product defended in vector spaces. The effectiveness of the proposed algorithms is tested by face recognition experiment results.

Abstract: The large dimensionality and unknown distributions are often met in a plant biotechnology and phytochemistry investigations. In this paper two methods are presented: principal component analysis allowing to reduce dimensionality and non-parametric Kruskal-Wallis ANOVA allowing to separate factors’ influence even if the distribution is unknown. The paper contains: problem definition, presentation of the measured data and the final analysis. The paper should be potentially useful to other industrial or research approaches.

Abstract: For the efficiency measurement of coal-fired power plant has drawn attention widely, and then this article analyzed five power plants in China with a combination of PCA and DEA. Adjust of production parameters and strategies according to the method to improve the efficiency of power plants. This article proved that the method can make a breakthrough in evaluating performance of the power plants.

Abstract: The T² statistic is one important indicator of statistical process control theory to identify anomalies of the multivariate industrial process. In the research field of the coal gas pre-drainage process control, previous achievements mainly based on the univariate control chart, which leaded to huge workload and facilitated some human errors. Against these problems, a more comprehensive and easy-to-use method based on the T² statistic was proposed. First at all, the basic thought and the principle of T² control chart was elaborated. Secondly, the data structure and data samples were provided after their principle component analysis. Finally, the multivariate control chart of coal gas pre-drainage process was established. Results show that the proposed anomaly identification method can integrate dozen of univariate control charts into one. Then technicians needn’t deal with many control charts in the same time and many human errors can be avoided.

Abstract: As one of the most popular and effective classification algorithms, Support Vector Machine (SVM) has attracted much attention in recent years. Classifiers ensemble is a research direction in machine learning and statistics, it often gives a higher classification accuracy than the single classifier. This paper proposes a new ensemble algorithm based on SVM. The proposed classification algorithm PB-SVM Ensemble consists of some SVM classifiers produced by PCAenSVM and fifty classifiers trained using Bagging, the results are combined to make the final decision on testing set using majority voting. The performance of PB-SVM Ensemble are evaluated on six datasets which are from UCI repository, Statlog or the famous research. The results of the experiment are compared with LibSVM, PCAenSVM and Bagging. PB-SVM Ensemble outperform other three algorithms in classification accuracy, and at the same time keep a higher confidence of accuracy than Bagging.

Abstract: The evaluation of PEG as a gelation solvent for an organogel based on β-cyclodextrin (β-CyD) was investigated using principal component analysis (PCA). The test tube tilting method was performed to examine the gel formation experimentally. The important descriptors used in the PCA included the Hansen partial solubility parameters. LogP may be able to be used as an additional descriptor for this system. According to PCA analysis, PEG was in a cluster of gelation solvents. Subsequently, various PEG liquid state grades were tested for their ability to gel the system and PEG400 was found to be able to produce a gel. This verifies that PCA can be successfully used to evaluate the role of PEG. Accordingly, this PCA method may be an effective tool to evaluate the role of any solvent for other low molecular weight gelators (LMWGs). Afterwards, the optimal components in the system were explored, and it was found to be 0.1M β-CyD, 0.5M K₂CO₃ in PEG400. Based on WAXS, the gel fiber of this organogel was demonstrated to be amorphous. This forming organogel will be further characterized in details and investigated for use in a drug delivery system.