Harald Steck

Title: Innovations of Harald Steck

Introduction

Harald Steck is an accomplished inventor based in New Providence, NJ (US). He has made significant contributions to the field of information processing and mobile localization. With a total of 2 patents, his work focuses on advanced systems that enhance user experience and device tracking.

Latest Patents

One of Harald Steck's latest patents is titled "Recommender system with training function based on non-random missing data." This invention involves a processing device that can obtain observed feedback data and construct a model that accounts for both observed and missing feedback data. The model is optimized using a training objective function to generate a list of recommended items for users. Notably, the missing feedback data is categorized as missing not at random (MNAR), and the model employs a matrix factorization approach.

Another significant patent is "KL-divergence kernel regression for non-gaussian fingerprint based localization." This invention addresses mobile localization and tracking of devices. It includes methods that utilize probability kernels with distance-like metrics between distributions. The probabilistic kernels described can be used for location regression, achieving impressive accuracy in office environments.

Career Highlights

Harald Steck is currently employed at Alcatel Lucent, where he continues to innovate and develop cutting-edge technologies. His work has had a profound impact on the fields of recommendation systems and mobile device tracking.

Collaborations

Throughout his career, Harald has collaborated with notable colleagues, including Piotr Mirowski and Philip Alfred Whiting. These collaborations have contributed to the advancement of his research and inventions.

Conclusion

Harald Steck's contributions to technology through his patents demonstrate his expertise and innovative spirit. His work in recommender systems and mobile localization continues to influence the industry and improve user experiences.