V
Viktor Losing
Researcher at Bielefeld University
Publications - 20
Citations - 669
Viktor Losing is an academic researcher from Bielefeld University. The author has contributed to research in topics: Data stream mining & Concept drift. The author has an hindex of 8, co-authored 16 publications receiving 443 citations. Previous affiliations of Viktor Losing include Honda.
Papers
More filters
Journal ArticleDOI
Incremental on-line learning: A review and comparison of state of the art algorithms
TL;DR: This work analyzes the key properties of eight popular incremental methods representing different algorithm classes and evaluates them with regards to their on-line classification error as well as to their behavior in the limit, facilitating the choice of the best method for a given application.
Proceedings ArticleDOI
KNN Classifier with Self Adjusting Memory for Heterogeneous Concept Drift
TL;DR: The Self Adjusting Memory (SAM) model for the k Nearest Neighbor (kNN) algorithm is proposed since kNN constitutes a proven classifier within the streaming setting and it can be easily applied in practice since an optimization of the meta parameters is not necessary.
Proceedings ArticleDOI
Interactive online learning for obstacle classification on a mobile robot
TL;DR: An architecture for incremental online learning in high-dimensional feature spaces and application on a mobile robot is presented, based on learning vector quantization, approaching the stability-plasticity problem of incremental learning by adaptive insertions of representative vectors.
Journal ArticleDOI
Tackling heterogeneous concept drift with the Self-Adjusting Memory (SAM)
TL;DR: This work proposes the Self-Adjusting Memory (SAM) model for the k-nearest-neighbor (kNN) algorithm, which can deal with heterogeneous concept drift, i.e., different drift types and rates, using biologically inspired memory models and their coordination.
Proceedings Article
Choosing the Best Algorithm for an Incremental On-line Learning Task
TL;DR: This work analyzes the key properties of seven incremental methods representing different algorithm classes and gives an overview of the performance with respect to accuracy as well as model complexity, facilitating the choice of the best method for a given application.