Fangfang Tan

Title: Innovations by Fangfang Tan in Predictive Modeling

Introduction

Fangfang Tan is an accomplished inventor based in Mountain View, CA. She has made significant contributions to the field of predictive modeling, particularly in the area of label shift detection and adjustment. Her innovative work has implications for improving the accuracy and reliability of machine learning models.

Latest Patents

Fangfang Tan holds a patent for "Label shift detection and adjustment in predictive modeling." This patent outlines techniques for detecting label shifts and adjusting training data of predictive models. In her invention, a first machine-learned model generates predicted labels for multiple scoring instances. The model is trained using various machine learning techniques based on a plurality of training instances, each containing an observed label. When a shift in observed labels is detected, the training data corresponding to specific segments is identified and adjusted. The adjusted training instances are then added to a final set of training data, which is used to train a second machine-learned model.

Career Highlights

Fangfang Tan is currently employed at Microsoft Technology Licensing, LLC, where she continues to develop her expertise in machine learning and predictive analytics. Her work is instrumental in advancing the capabilities of predictive modeling technologies.

Collaborations

Fangfang has collaborated with notable colleagues, including Jilei Yang and Yu Liu, who contribute to her innovative projects and research endeavors.

Conclusion

Fangfang Tan's contributions to predictive modeling through her patent on label shift detection and adjustment demonstrate her expertise and commitment to advancing technology. Her work at Microsoft Technology Licensing, LLC, along with her collaborations, positions her as a key figure in the field of machine learning.