by Vas Petridis, Ath Kehagias
http://users.auth.gr/~kehagiat/kehPub/journal/1996MAPPartitionTS.pdf
Add To MetaCart
Abstract:
e-maih kehagiasegnatia.ee.auth.gr We apply the Partition Algorithm to the problem of time series classification. We assume that the source that generates the time series belongs to a finite set of candidate sources. Classification is based on the computation of posterior probabilities. Prediction error is used to adaptively update the posterior probability of each source. The algorithm is implemented by a hierarchical, modular, recurrent network. The bottom (partition) level of the network consists of neural modules, each one trained to predict the output of one candidate source. The top (decision) level consists of a decision module, which computes posterior probabilities and classifies the time series to the source of maximum posterior probability. The classifier network is formed fi'om the composition of the partition and decision levels. This method applies to deterministic as well as probabilistic time series. Source switching can also be accommodated. We give some examples of application to problems of signal detection, phoneme and enzyme classification. In conclusion, the algorithm presented here gives a systematic method for the design of modular classification networks. The method can be extended by various choices of the partition and decision components.
Citations
|
1390
|
Introduction to the Theory of Neural Computation
– Hertz, Krogh, et al.
- 1991
|
|
569
|
Adaptive mixtures of local experts
– Jacobs, Jordan, et al.
- 1991
|
|
142
|
An Introduction to the Application of the Theory of Probabilistic Functions of a Markov Process to Automatic Speech Recognition
– Levinson, Rabiner, et al.
- 1983
|
|
96
|
A general regression neural network
– Specht
- 1991
|
|
75
|
Maximum likelihood competitive learning
– Nowlan
- 1990
|
|
66
|
Hierarchies of adaptive experts
– Jordan, Jacobs
- 1992
|
|
64
|
Hierarchical Mixtures of Experts and
– Jordan, Jacobs
- 1994
|
|
52
|
Equivalence Proofs for Multi-Layer Perceptron Classifiers and the Bayesian Discriminant Function
– Hampshire, Pearlmutter
- 1991
|
|
43
|
Estimation theory with applications to communications and control
– Sage, Melsa
- 1971
|
|
25
|
Neural network classification: A Bayesian interpretation
– Wan
- 1990
|
|
17
|
Competing Experts: An Experimental Investigation of Associative Mixture Models
– Nowlan
- 1990
|
|
16
|
Bayesian Mixture Modeling by Monte Carlo Simulation
– Neal
- 1991
|
|
13
|
Optimal Adaptive Estimation: Structure and Parameter Adaptation
– Lainiotis
- 1971
|
|
12
|
Bayes statistical behavior and valid generalization of pattern classifying neural networks
– Kanaya, Miyake
- 1991
|
|
12
|
Hidden control neural architecture modeling of nonlinear time varying systems and its applications
– Levin
- 1993
|
|
10
|
Recursive algorithm for the calculation of the adaptive Kalman filter coefficients
– Sims, Lainiotis, et al.
- 1969
|
|
8
|
Classification of Radar Clutter Using Neural Networks
– Haykin, Deng
- 1991
|
|
7
|
Learning stochastic feedforward networks
– Neal
- 1990
|
|
6
|
Optimal estimation in the presence of unknown parameters
– Hilborn, Lainiotis
- 1969
|
|
4
|
Adaptive deconvolution of seismic signals: performance, computational analysis, parallelism
– Lainiotis, Katsikas, et al.
- 1988
|
|
4
|
Maximum likelihood competition in RBF networks
– Nowlan
- 1990
|
|
3
|
Optimal adaptive filter realizations for stochastic processes with an unknown parameter
– Hilborn, Laintorts
- 1969
|
|
3
|
A neural network approach to a Bayesian statistical decision problem
– Miyake, Kanaya
- 1991
|
|
3
|
Adaptive Dynamic Neural Network Estimators
– Lainiotis, Plataniotis
- 1994
|
|
3
|
A Method for Bearings-Only Velocity and Position Estimation
– Petridis
- 1981
|
|
1
|
et al., "Task decomposition through competition in a modular connectionist architecture: the what and where vision tasks
– Jacobs
|
|
1
|
et al., "A Maximum Likelihood Approach to Continuous Speech Recognition
– Jelinek
|
|
1
|
Discrimination of Extended-Spectrum/%Lactamases by a Novel Nitrocefin Competition Assay
– Papanicolaou, Medeiros
- 1990
|
|
1
|
Phoneme recognition using time-delay neural networks
– eta
- 1989
|
|
1
|
Modularity and scaling in large phonemic neural networks
– eta
- 1989
|
|
1
|
et al., "Computer analysis of EEG signals with parametric models
– Zetterberg
- 1981
|
|
1
|
et al., "Training neural networks for ECG feature Recognition
– Zhu
- 1989
|
