M
Martin B. Scholz
Researcher at Hewlett-Packard
Publications - 37
Citations - 1913
Martin B. Scholz is an academic researcher from Hewlett-Packard. The author has contributed to research in topics: Collaborative filtering & Matrix (mathematics). The author has an hindex of 12, co-authored 37 publications receiving 1707 citations.
Papers
More filters
Proceedings ArticleDOI
One-Class Collaborative Filtering
TL;DR: This paper considers the one-class problem under the CF setting, and proposes two frameworks to tackle OCCF, one based on weighted low rank approximation; the other based on negative example sampling.
Journal ArticleDOI
Apples-to-apples in cross-validation studies: pitfalls in classifier performance measurement
George Forman,Martin B. Scholz +1 more
TL;DR: It is shown by experiment that all but one of these computation methods leads to biased measurements, especially under high class imbalance, which is of particular interest to those designing machine learning software libraries and researchers focused onhigh class imbalance.
Proceedings ArticleDOI
Mind the gaps: weighting the unknown in large-scale one-class collaborative filtering
Rong Pan,Martin B. Scholz +1 more
TL;DR: This paper proposes two novel algorithms for large-scale OCCF that allow to weight the unknowns: Low-rank matrix approximation, probabilistic latent semantic analysis, and maximum-margin matrix factorization.
Proceedings ArticleDOI
Feature shaping for linear SVM classifiers
TL;DR: This paper attempts to re-shape each input feature so that it is appropriate to use with a linear weight and to scale the different features in proportion to their predictive value, and demonstrates that this pre-processing is beneficial for linear SVM classifiers on a large benchmark of text classification tasks as well as UCI datasets.
Patent
Adjusting Content To User Profiles
TL;DR: One embodiment is a method that determines at a client computer a relevancy of information received with respect to a user profile and adjusts a ranking of the information according to the relevance and displays a selected portion of the adjusted information on the client computer as discussed by the authors.