INTERPRETATION OF CREDIT CARD FEATURES BY LIME AND SHAP AFTER DEEP LEARNING
Main Article Content
Abstract
Feature Interpretation after training is a necessary research trend in AI application and is of great interest to the AI research community. Explainable AI (XAI) has two main popular approaches, LIME and SHAP, which are very effective in explaining the models after training. There are many documents for using LIME and SHAP libraries to explain the models after training, but very few documents discuss the mathematical model inside LIME and SHAP. This makes it difficult to study. The paper describes in detail the steps of LIME and SHAP on small data and then explains its features after classifying credit card data by deep learning. The experimental results and interpretation show interesting information in interpreting the features by the LIME and SHAP methods.
Keywords
credit card clasification, LIME, machine learning, SHAP, XAI
Article Details
References
Chen, J., Song, L., Wainwright, M. J., & Jordan, M. I. (2018). Learning to explain: An information-theoretic perspective on model interpretation. In Proceedings of the 35th International Conference on Machine Learning (ICML) (pp. 883-892).
Dataset (2016). Default of Credit Card Clients. https://archive.ics.uci.edu/dataset/350/default+of+credit+card+clients
Dighe, D., Patil, S., & Kokate, S. (2018). Detection of credit card fraud transactions using machine learning algorithms and neural networks: a comparative study. In Proceedings of IEEE (pp.1-6).
Gu, L., Zhou, N., & Zhao, Y. (2018). An Euclidean Distance Based on Tensor Product Graph Diffusion Related Attribute Value Embedding for Nominal Data Clustering. Proceedings of the AAAI Conference on Artificial Intelligence, 32(1), 3101-3108. https://doi.org/10.1609/aaai.v32i1.11681
Guidotti, R., Monreale, A., Ruggieri, S., Turini, F., Giannotti, F., & Pedreschi, D. (2020). A Survey of Methods for Interpreting Machine Learning Models. ACM Computing Surveys, 52(5), 1-52.
Gramegna, A., & Giudici, P. (2021). SHAP and LIME: An Evaluation of Discriminative Power in Credit Risk. Frontiers in Artificial Intelligence, 4, Article 752558. https://doi.org/10.3389/frai.2021.752558
Kanmani, W. S., & Jayapradha, B. (2017). Prediction of default customer in banking sector using artificial neural network. International Journal of Research in Information Technology and Computing, 5, 293-296.
Lundberg, S. M., & Lee, S. I. (2017). Explaining black box models with Shapley values. In Advances in Neural Information Processing Systems (pp. 4765-4774).
Molnar, C. (2020). Interpretable Machine Learning. Leanpub. https://christophm.github.io/interpretable-ml-book/
Ribeiro, M. T., Singh, S., & Guestrin, C. (2017). A unified approach to interpretable machine learning. In Proceedings of the 34th International Conference on Machine Learning (ICML) (pp. 1165-1174).
Ribeiro, M. T., & Singh, S. (2020). "Why should I trust you?" Explaining the predictions of any machine learning classifier with LIME. Data Mining and Knowledge Discovery, 34(3), 818-835.
Ribeiro, M. T., Singh, S., & Guestrin, C. (2016). LIME: Local interpretable model-agnostic explanations. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD) (pp. 2145-2154).
Zhang, G. P. (2000). Neural networks for classification: A survey. IEEE Transactions on Systems, Man, and Cybernetics—Part C: Applications and Reviews, 30(4), 451-462.
Zhang, B., Wu, B., & Xu, L. (2022). Interpretable Machine Learning for Natural Language Processing. ACM Computing Surveys, 55(3), 1-40.