Manuel Nonnenmacher

Title: Manuel Nonnenmacher: Innovator in Neural Network Pruning

Introduction

Manuel Nonnenmacher is a notable inventor based in Holzgerlingen, Germany. He has made significant contributions to the field of computer-implemented neural networks. His innovative approach focuses on simplifying complex neural network structures, which can enhance computational efficiency.

Latest Patents

Nonnenmacher holds a patent titled "Efficient second order pruning of computer-implemented neural networks." This patent describes a method for generating a simplified computer-implemented neural network. The method involves receiving a predefined neural network, which includes various structures described by weights. Each structure is assigned a pruning vector that indicates changes in weights due to pruning. The process includes calculating a product of a matrix with partial second order derivations of a loss function concerning the weights. It determines changes in the loss function resulting from pruning and prunes at least one structure based on these changes to create a simplified neural network.

Career Highlights

Manuel Nonnenmacher is currently employed at Robert Bosch GmbH, where he applies his expertise in neural networks. His work is pivotal in advancing the efficiency of machine learning models. Nonnenmacher's innovative methods contribute to the ongoing evolution of artificial intelligence technologies.

Collaborations

He collaborates with esteemed colleagues, including David Reeb and Thomas Pfeil. Their combined efforts foster a creative environment that drives innovation in their projects.

Conclusion

Manuel Nonnenmacher's contributions to the field of neural networks exemplify the impact of innovative thinking in technology. His patent on efficient pruning methods showcases his commitment to enhancing computational efficiency in artificial intelligence.