<![CDATA[Public transport services are facing more challenges and some problems are gradually emerging with the increase in public transport users and varying travel demands. BRT as a type of public transportation in determining efficiency, service feasibility, and determining operational costs refers to the load factor. BRT bus load factor classification using KNN is proposed with parameter tuning to increase accuracy values. KNN algorithm with tuning parameters on 2 types of matrices (Minkowski and Euclidean) for BRT load factor classification for Transjatim corridor 1. The BRT load factor classification with the KNN algorithm increased by 7.81% by tuning parameters on the Euclidean matrix compared to the Minkowski matrix. The increase in accuracy is reflected in the confusion matrix with changes in increasing true negatives and decreasing false positives. Category 1 has a higher class than category 2 for boarding and alighting passengers. The classification presented can be a reference for Transjatim Corridor 1 Managers in determining efficiency, feasibility and operational costs.]]>
Copyrights © 2024
