Mathematics and Statistics Vol. 10(5), pp. 971 - 980
DOI: 10.13189/ms.2022.100508
Reprint (PDF) (345Kb)

Using Clustering Methods to Detect the Revealed Preferences of Moroccans towards the Electric Vehicles: Latent Class Analysis (LCA) and K-Modes Algorithm (K-MA)

Taoufiq El Harrouti ^1,*, Mourad Azhari ², Hajar Deqqaq ¹, Abdellah Abouabdellah ¹, Sanaa El Aidi ³, Habiba Chaoui ^3
1 Engineering Science Laboratory, National School of Applied Sciences, Ibn Tofail University, Kenitra, Morocco
² Center of Guidance and Planning of Education, MEN, Rabat, Morocco
³ Laboratory of Advanced Systems Engineering, National School of Applied Sciences, Ibn Tofail University, Kenitra, Morocco

ABSTRACT

Latent Class Analysis (LCA) and k-Mode Algorithm (K-MA) are two unsupervised machine learning techniques. These methods aim to identify individuals on the basis of their shared traits. They are utilized in the context of categorical data and can be used to detect people's opinions toward green forms of transportation, especially Electric Vehicles (EV) as an alternative to conventional internal combustion engine vehicles. The LCA approach discovers group profiles (clusters) based on observed variables, whereas the K-MA technique is an adaptation of the k-means algorithm for categorical variables. In this study, we apply these two methods to identify Moroccans' preferences for the electrification of their means of transportation. Both algorithms are able to divide the analyzed sample into two groups, with the first group being more interested in EV. The second group consists of individuals who are less concerned about ecologically sustainable transportation. In addition, we conclude that the LCA algorithm performs well and is superior to the K-MA, and that its discrimination power (65% vs 35%) is more than that of the K-MA (52% vs 48%).

KEYWORDS
Clustering, Latent Class Analysis (LCA), K-Mode Algorithm (K-MA), Electric Vehicles (EV), Survey

Cite This Paper in IEEE or APA Citation Styles
(a). IEEE Format:
[1] Taoufiq El Harrouti , Mourad Azhari , Hajar Deqqaq , Abdellah Abouabdellah , Sanaa El Aidi , Habiba Chaoui , "Using Clustering Methods to Detect the Revealed Preferences of Moroccans towards the Electric Vehicles: Latent Class Analysis (LCA) and K-Modes Algorithm (K-MA)," Mathematics and Statistics, Vol. 10, No. 5, pp. 971 - 980, 2022. DOI: 10.13189/ms.2022.100508.

(b). APA Format:
Taoufiq El Harrouti , Mourad Azhari , Hajar Deqqaq , Abdellah Abouabdellah , Sanaa El Aidi , Habiba Chaoui (2022). Using Clustering Methods to Detect the Revealed Preferences of Moroccans towards the Electric Vehicles: Latent Class Analysis (LCA) and K-Modes Algorithm (K-MA). Mathematics and Statistics, 10(5), 971 - 980. DOI: 10.13189/ms.2022.100508.