Sign in Subscribe
Machine learning

Artificial Intelligence Research Yields New Insights into Machine Learning

Strategic Insights

— 1 min read

Recent research at Dalian University of Technology in Liaoning, People's Republic of China, has made significant strides in the field of artificial intelligence, particularly in the area of machine learning. The study, published in the Journal of Engineering, has introduced a new framework for counterfactual explanation in machine learning, addressing the limitations of existing methods. This innovative approach enables users to generate more realistic and feasible counterfactual instances, enhancing the interpretability of machine learning models.

Key Takeaways:

  • The study proposes a new counterfactual explanation framework that introduces a feature-variable cost weight matrix, allowing for more realistic and feasible counterfactual results.
  • The defined objective function considers three indicators: feature-weighted distance, sparsity, and proximity, ensuring the feasibility, simplicity, and closeness to the original sample set of counterfactual results.
  • Genetic algorithms are used to solve the problem and generate the optimal action plan.
  • The research concludes that the proposed method can generate feasible and actionable counterfactual instances compared to existing counterfactual methods.
  • The study focuses on credit risk control scenarios, but the proposed framework can be applied to other fields as well.
  • The authors, WU Guowei and WANG Baocai, are affiliated with the School of Software Technology at Dalian University of Technology.
  • The research has the potential to improve the interpretability of machine learning models and provide users with more meaningful insights into model predictions.

Statistics:

  • 51 (2024) issue of Jisuanji kexue, a Chinese computer science journal, published the study.
  • The research was conducted at Dalian University of Technology in Liaoning, People's Republic of China.
  • The proposed framework is applicable to various fields, including finance, marketing, and healthcare.
  • The study used real datasets to test the efficacy of the proposed method.
  • The fault-tolerant rate of the proposed method is comparable to or better than existing counterfactual methods.

Sources:

  • Feature-weighted Counterfactual Explanation Method:A Case Study in Credit Risk Control Scenarios. Jisuanji kexue, 2024,51(12):259-268.
  • Editorial office of Computer Science (publisher)
  • WU Guowei, WANG Baocai (authors)
  • Dalian University of Technology, School of Software Technology (affiliation)
  • Doi: 10.11896/jsjkx.240300047 (link to journal article)

Read more