Trajectory classification-based car loan collection scoring model using a hierarchical CNN-LSTM method

Ming Pei

doi:10.1117/12.3009131

16 October 2023 Trajectory classification-based car loan collection scoring model using a hierarchical CNN-LSTM method

Ming Pei

Proceedings Volume 12803, Fifth International Conference on Artificial Intelligence and Computer Science (AICS 2023); 128033R (2023) https://doi.org/10.1117/12.3009131
Event: 2023 5th International Conference on Artificial Intelligence and Computer Science (AICS 2023), 2023, Wuhan, China

Abstract

To avoid economic losses on loan borrowers who are in default, financial institutions and companies employ plenty of debt collection measures like making phone calls, sending messages and even resorting to legal action. Since these measures also incur non-negligible costs and differ significantly in expenses, debt collection scoring models are proposed to assess the likelihood of a client’s payment delinquency thus helping companies in determining appropriate collection measures and improving effectiveness on allocating resources. Traditional methods in this area utilize clients’ personal information and are highly dependent on the accuracy of data which in practice can’t be guaranteed. In this work, we formulate collection scoring as a trajectory classification method and build a discriminative network that mines clients’ personal car trajectories which is an informative and stably accessible data resource in the car loan business scenario. We first propose a novel preprocessing method to extract features and transform raw trajectories into sequences of fixed-shape tensors. Then a convolutional auto-encoder is employed to condense extracted tensors into a low-dimensional representation. Finally, a Bi-LSTM is adopted to capture latent characteristics in the feature sequences and makes a prediction. Experiments on a real-world collected dataset confirm the effectiveness of our model.

Citation Download Citation

Ming Pei "Trajectory classification-based car loan collection scoring model using a hierarchical CNN-LSTM method", Proc. SPIE 12803, Fifth International Conference on Artificial Intelligence and Computer Science (AICS 2023), 128033R (16 October 2023); https://doi.org/10.1117/12.3009131

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
8 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Education and training

Feature extraction

Neural networks

Data modeling

Matrices

Global Positioning System

Mining

Show All Keywords

Keywords/Phrases

Search In:

Publication Years