Alan J Cossitt

Title: Alan J Cossitt: Innovator in Neural Network Training Systems

Introduction

Alan J Cossitt is a notable inventor based in Portland, OR (US). He has made significant contributions to the field of electronic networks, particularly in the training of neural networks. His innovative approach has led to the development of a unique training system that enhances the functionality of electronic networks.

Latest Patents

Cossitt holds a patent for a "Training system for neural networks and the like." This invention provides a means and method for training an electronic network of analog cells. The patent describes a generic universal programmer interface (GUPI) that enables the connection of an adaptor to a host computer via a personal computer personal programmer (PCPP) for training an ETANN chip. The adaptor plugs into the GUPI module, allowing users to connect either an ETANN chip or a Confidence Test Module (CTM). The invention includes various resources for training the ETANN, such as digital to analog conversion, neuron perturbation circuitry, and host computer software necessary for operation.

Career Highlights

Cossitt has had a distinguished career, working at Intel Corporation, where he has been able to apply his innovative ideas in a practical setting. His work has contributed to advancements in neural network technology, making significant impacts in the field.

Collaborations

Throughout his career, Cossitt has collaborated with notable colleagues, including Raymond E McNamee and Louis J Johnson. These collaborations have fostered an environment of innovation and have led to the development of groundbreaking technologies.

Conclusion

Alan J Cossitt's contributions to the field of neural networks exemplify the spirit of innovation. His patent for a training system showcases his ability to merge technology with practical applications, making a lasting impact in the industry.