• Publications
  • Influence
New Methods for Spectral Clustering.
Analyzing the affinity matrix spectrum is an increasingly popular data clustering method. We propose three new algorithmic components which are appropriate for improving performance of spectralExpand
  • 51
  • 5
  • PDF
Prediction with Expert Advice by Following the Perturbed Leader for General Weights
TLDR
We derive loss bounds for adaptive learning rate and both finite expert classes with uniform weights and countable Expert classes with arbitrary weights. Expand
  • 29
  • 5
  • PDF
Adaptive Online Prediction by Following the Perturbed Leader
TLDR
We derive loss bounds for adaptive learning rate and both finite expert classes with uniform weights and countable Expert classes with arbitrary weights. Expand
  • 75
  • 4
  • PDF
On the spectral mapping theorem for perturbed strongly continuous semigroups
Abstract. We consider a strongly continuous semigroup $(T(t))_{t \geqq 0}$ with generator A on a Banach space X, an A-bounded perturbation B, and the semigroup $(S(t))_{t \geqq 0}$ generated by A +Expand
  • 29
  • 4
The Critical Spectrum of a Strongly Continuous Semigroup
For a strongly continuous semigroup (T(t))t⩾0 with generator A we introduce its critical spectrum σcrit(T(t)). This yields in an optimal way the spectral mapping theorem σ(T(t))=etσ(A)∪σcrit(T(t))Expand
  • 36
  • 4
Amplifying the Block Matrix Structure for Spectral Clustering.
Spectral clustering methods perform well in cases where classical methods (K-means, single linkage, etc.) fail. However, for very non-compact clusters, they also tend to have problems. In this paper,Expand
  • 49
  • 3
  • PDF
BESS Control Strategies for Participating in Grid Frequency Regulation
Abstract Battery Energy Storage Systems (BESS) are very effective means of supporting system frequency by providing fast response to power imbalances in the grid. However, BESS are costly, andExpand
  • 56
  • 3
  • PDF
Main vector adaptation: a CMA variant with linear time and space complexity
TLDR
Adapting the main mutation vector instead of the covariance matrix yields an adaptation mechanism with space and time complexity O(N). Expand
  • 31
  • 3
  • PDF
Defensive Universal Learning with Experts
TLDR
This paper shows how universal learning can be achieved with expert advice. Expand
  • 22
  • 3
  • PDF
Memetic Algorithms For Combinatorial Optimization Problems In The Calibration Of Modern Combustion Engines
begin for j := 1 to do i := Local-Search(generateSolution()); add individual i to P ; endfor; repeat for i := 1 to ncross select two parents ia; ib 2 P randomly; ic := Local-Search(Recombine(ia;Expand
  • 14
  • 3
  • PDF
...
1
2
3
4
5
...