Masafumi Shingu

Title: Masafumi Shingu: Inventor in Model Generation and Machine Learning

Introduction

Masafumi Shingu is an inventor based in Mitaka, Japan, associated with Fujitsu Corporation. His patent work includes a method for generating a linear regression model using data processed by a machine-learning model.

Latest Patents

Shingu is credited with 1 patent: “Non-transitory Computer-Readable Recording Medium for Storing Model Generation Program, Model Generation Method, and Model Generation Device.” The invention describes generating multiple variations of input data, calculating distances between the original and generated data, and assessing how uniformly those distances are distributed. When the uniformity meets a specified threshold, the method generates a linear regression model using the machine-learning model’s outputs as the objective variable and the generated data as explanatory variables.

Career Highlights

Shingu’s listed patent reflects work at the intersection of machine learning, data analysis, and interpretable modeling. The approach uses a distribution-based condition to determine when to create a linear regression model from machine-learning results.

Collaborations

Shingu’s listed coworkers include Koji Maruhashi and Masaru Todoriki. His patent is associated with Fujitsu Corporation.

Conclusion

Masafumi Shingu’s patent describes a data-driven approach to generating linear regression models from machine-learning outputs, highlighting his contribution to model generation technology.