Abstract
Original language  English 

Qualification  Doctor Degree 
Awarding Institution 

Award date  20 Dec 1991 
Place of Publication  Espoo 
Publisher  
Print ISBNs  9513840611 
Publication status  Published  1991 
MoE publication type  G5 Doctoral dissertation (article) 
Fingerprint
Keywords
 neural networks
 classification
 nonlinear network analysis
Cite this
}
Time dependent classification of features with a nonlinear, dynamic network : Dissertation. / Vepsäläinen, Ari.
Espoo : VTT Technical Research Centre of Finland, 1991. 158 p.Research output: Thesis › Dissertation › Collection of Articles
TY  THES
T1  Time dependent classification of features with a nonlinear, dynamic network
T2  Dissertation
AU  Vepsäläinen, Ari
N1  Project code: TIK1007
PY  1991
Y1  1991
N2  Temporally varying classification by a dynamic classifier network is introduced. The dynamic classifier network consists of several independent nonlinear classifiers in parallel. The subclassifiers adapt to the measurements with a variety of adaptation rates. The output of the classifier network can be calculated as a weighted sum of the outputs of each subclassifier. Two methods to optimize the weighting are given. However, even a simple weighting function gives reasonable results. The network might be considered as a temporal associative memory. Because of nonlinearities and the ensuing chaos the behavior of the network can be very complicated. Algorithms to calculate the fractal and correlation dimension are also given. With these dimensions one can estimate how complicated the behavior of a system is and how many parameters are needed to describe its behavior. An extension of geodesic distance transform, called distance transform in curved space, is also presented. This transform can, for example, be used to model dynamic decision manifolds. Some new properties of fractals are also presented. These properties can be utilized efficiently when defining the Lyapunov exponents and the basins of attraction for maps. The methods presented have several application areas.
AB  Temporally varying classification by a dynamic classifier network is introduced. The dynamic classifier network consists of several independent nonlinear classifiers in parallel. The subclassifiers adapt to the measurements with a variety of adaptation rates. The output of the classifier network can be calculated as a weighted sum of the outputs of each subclassifier. Two methods to optimize the weighting are given. However, even a simple weighting function gives reasonable results. The network might be considered as a temporal associative memory. Because of nonlinearities and the ensuing chaos the behavior of the network can be very complicated. Algorithms to calculate the fractal and correlation dimension are also given. With these dimensions one can estimate how complicated the behavior of a system is and how many parameters are needed to describe its behavior. An extension of geodesic distance transform, called distance transform in curved space, is also presented. This transform can, for example, be used to model dynamic decision manifolds. Some new properties of fractals are also presented. These properties can be utilized efficiently when defining the Lyapunov exponents and the basins of attraction for maps. The methods presented have several application areas.
KW  neural networks
KW  classification
KW  nonlinear network analysis
M3  Dissertation
SN  9513840611
T3  Publications / Technical Research Centre of Finland
PB  VTT Technical Research Centre of Finland
CY  Espoo
ER 