Constructing a Hybrid Algorithm to Model the Physical and Chemical Inspection Station Data of the Shatt Al-Arab Waters*

Authors

  • Ahmed Husham Mohammed
  • Marwan Abdul Hameed Ashour

DOI:

https://doi.org/10.33095/qe4m1z69

Keywords:

Non-linear Data, nonparametric, KPCA, Fuzzy clustering, FCM.

Abstract

The aim of this paper is to find a hybrid method between of statistical methods that deal with non-linear high-dimensions. The data often suffer from complexity and overlap problems in their mathematical functions. It is difficult to delineate or accurately determine the effect of each variable on the other, accordingly it was constructing a hybrid model between the KPCA and FCM methods. The KPCA method aims to address the problem of high-dimensional nonlinear data and reduce it by finding a kernel matrix that depends primarily on the smoothing parameter matrix  that was estimated using the ROT method. Then, the FCM was adopted to obtain the clusters. This proposal was applied to the water sector in Basrah Governorate through a study of (8) stations for physical and chemical examination through (15) variables for the years (2019, 2020, 2021) and data were collected on a monthly basis. Through the application of this methodology, the paper was able to determine (7) Basic variables, which are (TH, Na, Cl, TDS, No3, EC, O_G). As for the stations, the overlapping stations between the clusters were identified, which are (SH1, SH2, SH3, SH4, E20, T34), as for the best degree of fuzziness it was (3.6) and the best number of clusters is (k = 3).

 

Downloads

Download data is not yet available.

Published

2024-07-18

Issue

Section

Statistical Researches

How to Cite

“Constructing a Hybrid Algorithm to Model the Physical and Chemical Inspection Station Data of the Shatt Al-Arab Waters*” (2024) Journal of Economics and Administrative Sciences, 30(141), pp. 457–478. doi:10.33095/qe4m1z69.

Similar Articles

1-10 of 953

You may also start an advanced similarity search for this article.