S
Simon Günter
Researcher at University of Bern
Publications - 32
Citations - 1683
Simon Günter is an academic researcher from University of Bern. The author has contributed to research in topics: Handwriting recognition & Hidden Markov model. The author has an hindex of 20, co-authored 32 publications receiving 1624 citations. Previous affiliations of Simon Günter include NICTA & Australian National University.
Papers
More filters
Proceedings Article
A stochastic quasi-Newton method for online convex optimization
TL;DR: Stochastic variants of the wellknown BFGS quasi-Newton optimization method, in both full and memory-limited (LBFGS) forms, are developed for online optimization of convex functions, which asymptotically outperforms previous stochastic gradient methods for parameter estimation in conditional random fields.
Journal ArticleDOI
Validation indices for graph clustering
Simon Günter,Horst Bunke +1 more
TL;DR: A new clustering algorithm for the domain of graphs is introduced and a number of cluster validation indices aim at finding the optimal number of clusters automatically.
Journal ArticleDOI
Fast Iterative Kernel Principal Component Analysis
TL;DR: Gain adaptation methods that improve convergence of the kernel Hebbian algorithm for iterative kernel PCA by incorporating the reciprocal of the current estimated eigenvalues as part of a gain vector in reproducing kernel Hilbert space to further speed up convergence.
Journal ArticleDOI
Self-organizing map for clustering in the graph domain
Simon Günter,Horst Bunke +1 more
TL;DR: This paper presents a version of self-organizing map that works in the domain of graphs and demonstrates its feasibility in the context of a graph clustering task.
Journal ArticleDOI
A Quasi-Newton Approach to Nonsmooth Convex Optimization Problems in Machine Learning
TL;DR: A new, efficient, exact line search algorithm that is comparable to or better than specialized state-of-the-art solvers on a number of publicly available data sets and proves its worst-case time complexity bounds.