Alexey Miroshnikov

Title: Alexey Miroshnikov: Innovator in Machine Learning Interpretability

Introduction

Alexey Miroshnikov is a notable inventor based in Evanston, IL, who has made significant contributions to the field of machine learning interpretability. With a total of three patents to his name, Miroshnikov's work focuses on developing frameworks that enhance the understanding of machine learning models.

Latest Patents

Miroshnikov's latest patents include a system and method for utilizing grouped partial dependence plots and game-theoretic concepts in the generation of adverse action reason codes. This framework proposes a method for interpreting machine learning models by determining the contribution of groups of input variables to the model's output. The input variables are grouped based on their dependencies with other variables, identified through a clustering algorithm applied to a training data set. Scores related to each group of input variables are calculated for specific instances of input vectors processed by the model. The algorithms employed can utilize group Partial Dependence Plot (PDP) values, Shapley Additive Explanations (SHAP) values, and Banzhaf values, among others. These scores can then be sorted, ranked, and combined into a hybrid ranking. Another patent focuses on a similar framework that utilizes grouped partial dependence plots and SHAP values for generating adverse action reason codes, emphasizing the importance of interpretability in machine learning.

Career Highlights

Miroshnikov is currently employed at Discover Financial Services LLC, where he applies his expertise in machine learning and data analysis. His innovative approaches have contributed to advancements in the financial services sector, particularly in enhancing the interpretability of complex models.

Collaborations

Some of Miroshnikov's notable coworkers include Konstandinos Kotsiopoulos and Arjun Ravi Kannan, who collaborate with him on various projects related to machine learning and data science.

Conclusion

Alexey Miroshnikov's work in machine learning interpretability showcases his commitment to advancing technology in a way that enhances understanding and transparency. His contributions are paving the way for more interpretable and reliable machine learning applications.