Study of taxi availability in Singapore using principal component analysis

• Main Author: Sonia Gunawan
• Other Authors: Zhu Feng
• Format: Final Year Project
• Language: English
• Published: 2018
• Subjects: DRNTU::Engineering::Civil engineering::Transportation
• Online Access: http://hdl.handle.net/10356/74669
• Institution: Nanyang Technological University

Due to the technology-driven market disruption caused by the introduction of Private Hire Car services in Singapore, conventional taxi services sustained a noticeable decrease in the number of fleets throughout the year. One of the main differences which differentiates these two services is the waiting time, which is mostly determined by their availability. Therefore, to keep up with the competition, the need to analyze taxi availability in Singapore arises. In this project, three types of dataset – taxi availability; speed bands and ERP rates, were extracted from LTA DataMall to be compared with each other. Data visualization using Google Maps API and statistical analysis of data were carried out to find the correlations of the collected datasets. Finally, a machine learning method using Principal Component Analysis was adopted to further analyze the extracted datasets. The extracted datasets were also analyzed by a machine learning method using Principal Component Analysis. It was found that PCA was not the most suitable method for the analysis of these dataset which was mainly due to the distribution of the dataset. Therefore, several approaches which would be more suitable were recommended under the conclusion, including Kernel PCA and Independent Component Analysis.