Frederick E Daum

Title: Innovations of Frederick E Daum

Introduction

Frederick E Daum is an accomplished inventor based in Carlisle, MA (US). He has made significant contributions to the field of neural networks, particularly through his innovative patent related to Bayesian neural networks.

Latest Patents

Frederick E Daum holds a patent for "Particle flow training of Bayesian neural network." This patent discusses devices, systems, and methods for training and operating a Bayesian neural network (BNN). The method includes initializing particles that represent pointwise values of respective neural network parameters. These parameters collectively represent a distribution of parameters of the BNN. The process involves optimizing the particles using training particle flow, resulting in optimized distributions for the parameters. Additionally, it determines a prediction distribution using the optimized distributions and predictions from each of the neural networks, ultimately providing a marginalized distribution representative of the prediction distribution.

Career Highlights

Frederick E Daum is associated with Raytheon Company, where he applies his expertise in neural networks and machine learning. His work has contributed to advancements in technology and innovation within the company.

Collaborations

Some of his notable coworkers include Suzanne M Baker and Andrew C Allerdt, who collaborate with him on various projects and initiatives.

Conclusion

Frederick E Daum's innovative work in the field of Bayesian neural networks exemplifies the impact of his contributions to technology. His patent reflects a significant advancement in the training and operation of neural networks, showcasing his expertise and dedication to innovation.