Causality is often measured by means of the correlation function. This function has been shown to be restrictive at best and false at worst. The accurate determination of causal relationships is part of the content of this work. To measure associations, strategies and methods other than simple correlation are required. In the first part, signals of differing natures are translated into a common domain, known in pattern recognition as feature space. This allows the determination of association over a large number of signals. In the second part, states within the mapping space are determined using clustering methods to form symbol sequences. In the last part of the thesis, interactions are determined using both symbol sequences and unsegmented feature vectors. Interactions are estimated based on the correlation function and directed trans-information, which is a generalisation of the correlation measure. All procedures are applied not only to synthetic data but also to signal recordings in a wide range of physiological conditions, such as CheyneStokes breathing and anaesthesia experiments. It is shown that Information Theory
|
4388
|
Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference
– Pearl
- 1988
|
|
4364
|
Elements of Information Theory
– Cover, Thomas
- 1991
|
|
4345
|
Maximum likelihood from incomplete data via the EM algorithm
– Dempster, Laird, et al.
- 1977
|
|
3052
|
Neural Networks for Pattern Recognition
– Bishop
- 1995
|
|
2961
|
Pattern Classification and Scene Analysis
– Duba, Hart
- 1973
|
|
1664
|
Introduction to statistical pattern recognition, 2nd ed
– Fukunaga
- 1990
|
|
938
|
Density Estimation for Statistics and Data Analysis
– Silverman
- 1986
|
|
699
|
Modeling by shortest data description
– Rissanen
- 1978
|
|
696
|
Time Series Analysis: Forecasting and Control
– Box, Jenkins, et al.
- 1994
|
|
688
|
An information-maximization approach to blind separation and blind deconvolution
– Bell, Sejnowski
- 1995
|
|
611
|
A new look at the statistical model identification
– Akaike
- 1974
|
|
589
|
Information Theory and Statistics
– Kullback
- 1959
|
|
257
|
Spectral Analysis and Time Series
– Priestley
- 1981
|
|
201
|
Linear prediction: A tutorial review
– Makhoul
- 1975
|
|
145
|
Unsupervised Optimal Fuzzy Clustering
– Gath, Geva
- 1989
|
|
119
|
Digital Spectral Analysis with Applications. Engineering
– Marple
- 1987
|
|
118
|
The Estimation of the Gradient of a Density Function, with Applications in Pattern Recognition
– Fukunaga, Hostetler
|
|
106
|
Probability theory
– Rényi
- 1970
|
|
100
|
Maximum likelihood and covariant algorithms for independent component analysis
– MacKay
- 1996
|
|
93
|
independent coordinates for strange attractors from mutual information, Phys
– Farmer, Swinney
- 1986
|
|
80
|
Information Flow in Non deterministic Systems
– Wittbold, Johnson
- 1990
|
|
66
|
Numerical Bayesian Methods Applied to Signal Processing
– Ruanaidh, Fitzgerald
- 1996
|
|
58
|
Extracting qualitative dynamics from experimental data
– Broomhead, King
- 1986
|
|
53
|
Spectrum analysis - a modern perspective
– Kay, SJ
- 1981
|
|
52
|
Coding and Information Theory
– Roman
- 1992
|
|
44
|
Electrical Fields of the Brain: The Neurophysics of EEG
– Nunez
- 1981
|
|
37
|
Mutual information functions versus correlation functions
– Li
- 1990
|
|
29
|
Introduction to Metric and Topological Spaces
– Sutherland
- 1975
|
|
27
|
On the Logic of Information Flow
– Barwise, Seligman
- 1995
|
|
24
|
Stochastic Complexity Measures for Physiological Signal Analysis
– Rezek, SJ
- 1998
|
|
21
|
Information and entropy in strange attractors
– Fraser
- 1989
|
|
20
|
Introduction to Minimum Encoding Inference
– Oliver, Hand
- 1994
|
|
17
|
M: "Approximate entropy as a measure of system complexity
– Pincus
- 1991
|
|
13
|
Reconstructing attractors from scalar time series: a comparison of singular system and redundancy criteria
– Fraser
- 1989
|
|
12
|
Chaos and Nonlinear Dynamics
– Hilborn
- 1994
|
|
10
|
Estimation of mutual information using kernel density estimators
– Moon, Rajagopalan, et al.
- 1995
|
|
9
|
Inconsistency of the AIC Rule for Estimating the Order of Autoregressive Models
– Kashyap
- 1980
|
|
8
|
Model order selection for the singular value decomposition and the discrete Karhunen-Loeve transform using a Bayesian approach
– Rajan, Rayner
- 1997
|
|
8
|
Physiologie des Menschen
– Schmidt, Thews
- 1997
|
|
7
|
On finite sample theory for autoregressive model order selection
– Broersen, Wensink
- 1993
|
|
7
|
Measuring correlations in symbol sequences
– Herzel, Gro��e
- 1995
|
|
6
|
Causal cognition: A multidisciplinary debate
– Sperber, Premack, et al.
- 1995
|
|
4
|
A correlation function for choosing time delays in phase portrait reconstructions
– Kember, Fowler
- 1993
|
|
4
|
Neural networks and nonlinear regression modelling and control of depth of anaesthesia for spontaneously breathing and ventilated patients. Computer Methods and Programming
– Rehman, Linkens, et al.
- 1993
|
|
3
|
A new method of the description of the information flow in the brain structures
– Kamifiski, Blinowska
- 1991
|
|
3
|
A time-series analysis method based on the directed transinformation
– Kamitake, Harashima, et al.
- 1984
|
|
3
|
Causality and information flow
– Lloyd
- 1991
|
|
2
|
Adaptive measurement of linear dependence between physiological signals by means of time-varying multivariate ar-modelling. submitted to
– Arnold, Miltner, et al.
- 1996
|
|
2
|
Autonomic Failure, A Textbook of Clinical Disorders of the Autonomic Nervous System
– Bannister, Mathias
- 1993
|
|
2
|
Biomedical Signal Processing : Time and Frequency domains analysis, volume 1
– Cohen
- 1986
|
