Online Principal Component Analysis in High Dimension: Which Algorithm to Choose?

To solve the problem of water management subsystem fault diagnosis in a proton exchange membrane fuel cell (PEMFC) system, a novel approach based on learning vector quantization neural network (LVQNN) and kernel principal component analysis (KPCA) is proposed. In the proposed approach, the KPCA method is used for processing strongly coupled fault data with a high dimension to reduce the data dimension and to extract new low-dimensional fault feature data. The LVQNN method is used to carry out fault recognition using the fault feature data. The effectiveness of the proposed fault detection method is validated using the experimental data of the PEMFC power system. Results show that the proposed method can quickly and accurately diagnose the three health states: normal state, water flooding failure and membrane dry failure, and the recognition accuracy can reach 96.93%. Therefore, the method proposed in this paper is suitable for processing the fault data with a high dimension and abundant quantities, and provides a reference for the application of water management subsystem fault diagnosis of PEMFC.

Download Full-text

A modified correlation in principal component analysis for torrential rainfall patterns identification

IAES International Journal of Artificial Intelligence (IJ-AI) ◽

10.11591/ijai.v9.i4.pp655-661 ◽

2020 ◽

Vol 9 (4) ◽

pp. 655

Author(s):

Shazlyn Milleana Shaharudin ◽

Norhaiza Ahmad ◽

Siti Mariana Che Mat Nor

Keyword(s):

Principal Component Analysis ◽

High Dimension ◽

Pearson Correlation ◽

Principal Component ◽

Breakdown Point ◽

Distribution Patterns ◽

Component Analysis ◽

Peninsular Malaysia ◽

Torrential Rainfall ◽

Rainfall Patterns

This paper presents a modified correlation in principal component analysis (PCA) for selection number of clusters in identifying rainfall patterns. The approach of a clustering as guided by PCA is extensively employed in data with high dimension especially in identifying the spatial distribution patterns of daily torrential rainfall. Typically, a common method of identifying rainfall patterns for climatological investigation employed T mode-based Pearson correlation matrix to extract the relative variance retained. However, the data of rainfall in Peninsular Malaysia involved skewed observations in the direction of higher values with pure tendencies of values that are positive. Therefore, using Pearson correlation which was basing on PCA on rainfall set of data has the potentioal to influence the partitions of cluster as well as producing exceptionally clusters that are eneven in a space with high dimension. For current research, to resolve the unbalanced clusters challenge regarding the patterns of rainfall caused by the skewed character of the data, a robust dimension reduction method in PCA was employed. Thus, it led to the introduction of a robust measure in PCA with Tukey’s biweight correlation to downweigh observations along with the optimal breakdown point to obtain PCA’s quantity of components. Outcomes of this study displayed a highly substantial progress for the robust PCA, contrasting with the PCA-based Pearson correlation in respects to the average amount of acquired clusters and indicated 70% variance cumulative percentage at the breakdown point of 0.4.

Download Full-text

Clustering by principal component analysis with Gaussian kernel in high-dimension, low-sample-size settings

Journal of Multivariate Analysis ◽

10.1016/j.jmva.2021.104779 ◽

2021 ◽

pp. 104779

Author(s):

Yugo Nakayama ◽

Kazuyoshi Yata ◽

Makoto Aoshima

Keyword(s):

Principal Component Analysis ◽

Sample Size ◽

High Dimension ◽

Principal Component ◽

Component Analysis ◽

Gaussian Kernel

Download Full-text

Rice Odours’ Readings Investigation Using Principal Component Analysis

International Journal of Engineering & Technology ◽

10.14419/ijet.v7i2.29.13803 ◽

2018 ◽

Vol 7 (2.29) ◽

pp. 488

Author(s):

Nurul Aini Abdul Wahab ◽

Shamshuritawati Sharif

Keyword(s):

Principal Component Analysis ◽

High Dimension ◽

Electronic Nose ◽

Principal Component ◽

Component Analysis ◽

The Other ◽

Total Variance ◽

Scree Plot ◽

Cumulative Proportion ◽

Principle Component

The use of electronic nose (e-nose) devices plus principal component analysis can help the process of categorizing the 16 different rice into its type. Generally, the physical feature of an e-nose own more than one hole to capture the odour of rice. For example, the portable e-nose so-called Insniff does have 10 holes (or variables). In this situations, we will have a dataset that consist high-dimension dataset where lead to the presence of interdependencies between all variables under study. Therefore, this study is presented to investigate the odour of rice for identifying the most important variables contributing to the rice odour readings. The principal component analysis (PCA) is implemented to determine the component that best represent the all 10 variables in order to eliminate the interdependency problem, and (2) to identify which variable is considered as important and influential to the newly-formed principle component (PC). The results from PCA suggested that the first two principle components is chosen. It is based on three assessments which are Kaiser’s criterion larger than 1, cumulative proportion of total variance, and scree plot. These two principle components explained 89% of total variance. Results showed that sensor 1 (0.931) and sensor 2 (0.966) are the two important variables that highly contribute to PC1. On the other hand, for PC2, the highest contribution is from sensor 8 (0.828). This study demonstrate that PCA is effective for investigating rice odour readings.

Download Full-text

A German version of the Intermittent Claudication Questionnaire (ICQ): cultural adaptation and validation

VASA ◽

10.1024/0301-1526/a000218 ◽

2012 ◽

Vol 41 (5) ◽

pp. 333-342 ◽

Cited By ~ 3

Author(s):

Kirchberger ◽

Finger ◽

Müller-Bühl

Keyword(s):

Principal Component Analysis ◽

Intermittent Claudication ◽

Completion Time ◽

Short Form ◽

Principal Component ◽

Component Analysis ◽

German Version ◽

Average Completion Time ◽

Sf 36 ◽

Related Quality

Background: The Intermittent Claudication Questionnaire (ICQ) is a short questionnaire for the assessment of health-related quality of life (HRQOL) in patients with intermittent claudication (IC). The objective of this study was to translate the ICQ into German and to investigate the psychometric properties of the German ICQ version in patients with IC. Patients and methods: The original English version was translated using a forward-backward method. The resulting German version was reviewed by the author of the original version and an experienced clinician. Finally, it was tested for clarity with 5 German patients with IC. A sample of 81 patients were administered the German ICQ. The sample consisted of 58.0 % male patients with a median age of 71 years and a median IC duration of 36 months. Test of feasibility included completeness of questionnaires, completion time, and ratings of clarity, length and relevance. Reliability was assessed through a retest in 13 patients at 14 days, and analysis of Cronbachs alpha for internal consistency. Construct validity was investigated using principal component analysis. Concurrent validity was assessed by correlating the ICQ scores with the Short Form 36 Health Survey (SF-36) as well as clinical measures. Results: The ICQ was completely filled in by 73 subjects (90.1 %) with an average completion time of 6.3 minutes. Cronbachs alpha coefficient reached 0.75. Intra-class correlation for test-retest reliability was r = 0.88. Principal component analysis resulted in a 3 factor solution. The first factor explained 51.5 of the total variation and all items had loadings of at least 0.65 on it. The ICQ was significantly associated with the SF-36 and treadmill-walking distances whereas no association was found for resting ABPI. Conclusions: The German version of the ICQ demonstrated good feasibility, satisfactory reliability and good validity. Responsiveness should be investigated in further validation studies.

Download Full-text