DOI : https://doi.org/10.18517/ijaseit.11.5.13429

The Monitoring of Dirichlet Compositional Data

Reham W. Elshaer (1), Aya A. Aly (2)
(1) Statistics Department, Cairo University, 1 Gamaa Street, Giza; 12613, Egypt
(2) Statistics Department, Cairo University, 1 Gamaa Street, Giza; 12613, Egypt
Fulltext View | Download
How to cite (IJASEIT) :
Elshaer, Reham W., and Aya A. Aly. “The Monitoring of Dirichlet Compositional Data”. International Journal on Advanced Science, Engineering and Information Technology, vol. 11, no. 5, Oct. 2021, pp. 1868-75, doi:10.18517/ijaseit.11.5.13429.
Citation Format :
Compositional data are used in many applications such as Cement, Asphalt, and many other Chemical industries. Such data represent random variables whose values must sum up to a certain constant. Quality engineers and technicians require monitoring compositional data and detecting the source of the irregularity in the process as soon as it happens. Throughout the literature, complicated methods were introduced to monitor compositional data. Such methods are computationally complex and can lead to difficulties in interpreting the results. The Dirichlet distribution is commonly used in the literature to model compositional data. In this study, we propose three simple methods to monitor the mean vector of the Dirichlet distribution. The first method is based on a MEWMA control chart. The second method is based on transforming the Dirichlet random variables into beta random variables and then monitoring them using multiple EWMA control charts, while the third method uses multiple EWMA control charts for transformed independent random variables. Using a simulation technique, the performance of the three methods is investigated, and the three methods performed very well under different sample sizes, many random variables, and values of the distribution parameters. When the process is out-of-control, the source of the out-of-control signal can be detected using Method 2 and Method 3. Method 2 maintained its good performance with a probability 0.99 of correctly detecting the source of the signal. Method 3 performed well except for the case of Dirichlet parameter values less than one. However, it maintained almost a probability of correct detection of at least 90% in most cases. The three proposed methods are simple, do not need complicated calculations, and can easily be applied and used by practitioners.

D. C. Montgomery, Introduction to Statistical Quality Control, Sixth Edition. 2009.

L. Foley, D. Dumuid, A. J. Atkin, T. Olds, and D. Ogilvie, “Patterns of health behaviour associated with active travel: A compositional data analysis,” Int. J. Behav. Nutr. Phys. Act., 2018, doi: 10.1186/s12966-018-0662-8.

T. P. Quinn, I. Erb, G. Gloor, C. Notredame, M. F. Richardson, and T. M. Crowley, “A field guide for the compositional analysis of any-omics data,” Gigascience, 2019, doi: 10.1093/gigascience/giz107.

K. Pearson, “Mathematical contributions to the theory of evolution. —On a form of spurious correlation which may arise when indices are used in the measurement of organs,” Proc. R. Soc. London, vol. 60, no. 359-367, pp. 489-498, Dec. 1897, doi: 10.1098/rspl.1896.0076.

V. Pawlowsky-Glahn and A. Buccianti, Compositional Data Analysis: Theory and Applications. 2011.

J. Aitchison, The Statistical Analysis of Compositional Data. 1986.

P. Praus, “Robust multivariate analysis of compositional data of treated wastewaters,” Environ. Earth Sci., 2019, doi: 10.1007/s12665-019-8248-6.

J. J. Egozcue, V. Pawlowsky-Glahn, and G. B. Gloor, “Linear association in compositional data analysis,” Austrian J. Stat., 2018, doi: 10.17713/ajs.v47i1.689.

V. Pawlowsky-Glahn, “Peter Filzmoser, Karel Hron, Matthias Templ: Applied compositional data analysis, with worked examples in R,” Stat. Pap., vol. 61, no. 2, pp. 921-922, 2020, doi: 10.1007/s00362-020-01163-7.

R. A. Boyles, “Using the chi-square statistic to monitor compositional process data,” J. Appl. Stat., 1997, doi: 10.1080/02664769723567.

G. Yang, D. B. H. Cline, R. L. Lytton, and D. N. Little, “Ternary and Multivariate Quality Control Charts of Aggregate Gradation for Hot Mix Asphalt,” J. Mater. Civ. Eng., 2004, doi: 10.1061/(asce)0899-1561(2004)16:1(28).

M. Vives-Mestres, J. Daunis-I-Estadella, and J. A. Martí­n-Ferní¡ndez, “Individual T2 control chart for compositional data,” J. Qual. Technol., 2014, doi: 10.1080/00224065.2014.11917958.

J. J. Egozcue, V. Pawlowsky-Glahn, G. Mateu-Figueras, and C. Barceló-Vidal, “Isometric Logratio Transformations for Compositional Data Analysis,” Math. Geol., 2003, doi: 10.1023/A:1023818214614.

M. Vives-Mestres, J. Daunis-I-Estadella, and J. A. Martí­n-Ferní¡ndez, “Out-of-control signals in three-part compositional T2 control chart,” 2014, doi: 10.1002/qre.1583.

M. Vives-Mestres, J. Daunis-i-Estadella, and J. A. Martí­n-Ferní¡ndez, “Signal interpretation in Hotelling’s T2 control chart for compositional data,” IIE Trans. (Institute Ind. Eng., 2016, doi: 10.1080/0740817X.2015.1125042.

K. P. Tran, P. Castagliola, G. Celano, and M. B. C. Khoo, “Monitoring compositional data using multivariate exponentially weighted moving average scheme,” Qual. Reliab. Eng. Int., 2018, doi: 10.1002/qre.2260.

F. Alt and K. Jain, “Multivariate quality controlMultivariate quality control,” in Encyclopedia of Operations Research and Management Science, S. I. Gass and C. M. Harris, Eds. New York, NY: Springer US, 2001, pp. 544-550.

A. Ongaro and S. Migliorati, “A generalization of the dirichlet distribution,” J. Multivar. Anal., 2013, doi: 10.1016/j.jmva.2012.07.007.

Authors who publish with this journal agree to the following terms:

    1. Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
    2. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
    3. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).