Tianqi Wan

Title: Innovations of Tianqi Wan in Neural Network Testing

Introduction

Tianqi Wan is an accomplished inventor based in Beijing, China. He has made significant contributions to the field of neural networks, particularly in assessing their test adequacy. His innovative approach combines various testing methodologies to enhance the reliability of deep learning systems.

Latest Patents

Tianqi Wan holds a patent titled "Method for assessing test adequacy of neural network based on element decomposition." This patent presents a method for evaluating the test adequacy of deep neural networks by decomposing key elements. The process involves dividing network testing into black box and white box testing, extracting essential parameters such as weight matrices and bias vectors. Furthermore, it calculates the importance values of neurons in individual layers and generates a hot map based on clustering results. The method also incorporates mutation testing and index calculation for comprehensive evaluation.

Career Highlights

Tianqi Wan is affiliated with the Beijing Aerospace Institute for Metrology and Measurement Technology. His work at this prestigious institution allows him to apply his innovative ideas in a practical setting, contributing to advancements in metrology and measurement technology.

Collaborations

Tianqi has collaborated with notable colleagues, including Yinxiao Miao and Yifei Liu. Their combined expertise fosters a productive environment for innovation and research.

Conclusion

Tianqi Wan's contributions to neural network testing exemplify the intersection of innovation and technology. His patent and work at the Beijing Aerospace Institute highlight his commitment to advancing the field.