A Comprehensive Analysis of Clustering Public Utility Bus Passenger’s Behavior during the COVID-19 Pandemic: Utilization of Machine Learning with Metaheuristic Algorithm

Abstract

Public utility bus (PUB) systems and passenger behaviors drastically changed during the COVID-19 pandemic. This study assessed the clustered behavior of 505 PUB passengers using feature selection, K-means clustering, and particle swarm optimization (PSO). The wrapper method was seen to be the best among the six feature selection techniques through recursive feature selection with a 90% training set and a 10% testing set. It was revealed that this technique produced 26 optimal feature subsets. These features were then fed into K-means clustering and PSO to find PUB passengers’ clusters. The algorithm was tested using 12 different parameter settings to find the best outcome. As a result, the optimal parameter combination produced 23 clusters. Utilizing the Pareto analysis, the study only considered the vital clusters. Specifically, five vital clusters were found to have comprehensive similarities in demographics and feature responses. The PUB stakeholders could use the cluster findings as a benchmark to improve the current system.

Indonesian martial artists’ preferences in martial arts schools: Sustaining business competitiveness through conjoint analysis

“…When participants submitted their responses one at a time, the Kolmogorov-Smirnov test and univariate analysis were continuously applied. These tests supported the significance of a sample size greater than 50 and data with underlying levels for each attribute [3,25]. Through IBM SPSS 25, the p-value was found statistically significant because all the groupings were less than 0.05.…”

Exploring Employee Retention among Generation Z Engineers in the Philippines Using Machine Learning Techniques

“…The wrapper method's backward elimination explores the optimal subset by feeding all features into the model and eliminating the least important feature one at a time [7]. Among all wrapper methods, many studies supported that its extensive process leads to a high accuracy value [23]. Backward elimination applies the Ordinary Least Squares model to identify the best feature combination depending on the class.…”

“…Backward elimination applies the Ordinary Least Squares model to identify the best feature combination depending on the class. This process is repeated continuously until the combination's p-values are less than or equal to 0.05 [23]. The p-value of the best feature combination was selected by adopting the equation from the study of Maldonado and Weber [24]:…”

“…As shown in Table 5, the most optimal feature selection was the wrapper method through backward elimination with 85.66% accuracy with a total of 15 important subfeatures. It was chosen because it garnered the highest optimal number with the highest accuracy [23]. According to a past study, the wrapper's backward elimination displays a higher accuracy value through continuous computations of group dependencies, resulting in minimized error [7].…”

Determinants of continuous visiting behavior to Palawan, Philippines: Integrating Uncertainty Reduction Theory and Expectation Confirmation Theory

“…Hence, employees must show professionalism and be well-equipped with training and knowledge to develop a positive service experience [ 13 ]. Based on past studies, the attitude of employees directly influenced service experience [ 13 , 28 , 29 ]. Through the relevant research, this study hypothesized that:…”