As data becomes more complex and more powerful deep learning architectures are used, the methods used to reach a solution become much less interpretable on the surface level. In recognition of such an issue, more developers are starting to incorporate explainable AI into their complex solutions in order to make these problems more understandable. This places more emphasis on human understanding, which is certainly crucial to a problem due to the degree of complexity that the most powerful solutions introduce. Furthermore, our research aims showcase the application of explainable AI to a complex and hard-to-interpret solution in the urban intelligence domain, specifically related to traffic data. In this paper, we build a deep classification model with the goal of predicting the identity of a driver responsible for a taxi trip, train it on complex data, and use explainable AI to better interpret the results of this model.

• This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator

• Kim, Jin Ryoul
• Coolidge, Ian
• Burke, William

Publisher

• Worcester Polytechnic Institute

Identifier

• 95176
• E-project-032423-131432

Advisor

• Li, Yanhua

Year

• 2023

Date created

• 2023-03-24

Resource type

• Major Qualifying Project

Major

• Computer Science

Source

• E-project-032423-131432

Rights statement

• In Copyright

Last modified

• 2023-04-19