Abstract
Clustering is a fundamental data processing technique. While clustering of static (vector based) data and of fixed window size time series have been well explored, dynamic clustering of spatiotemporal data has been little researched if at all. Especially when patterns of changes (events) in the data across space and time have to be captured and understood. The paper presents novel methods for clustering of spatiotemporal data using the NeuCube spiking neural network (SNN) architecture. Clusters of spatiotemporal data were created and modified on-line in a continuous, incremental way, where spatiotemporal relationships of changes in variables are incrementally learned in a 3D SNN model and the model connectivity and spiking activity are incrementally clustered. Two clustering methods were proposed for SNN, one performed during unsupervised and one—during supervised learning models. Before submitted to the models, the data is encoded as spike trains, a spike representing a change in the variable value (an event). During the unsupervised learning, the cluster centres were predefined by the spatial locations of the input data variables in a 3D SNN model. Then clusters are evolving during the learning, i.e. they are adapted continuously over time reflecting the dynamics of the changes in the data. In the supervised learning, clusters represent the dynamic sequence of neuron spiking activities in a trained SNN model, specific for a particular class of data or for an individual instance. We illustrate the proposed clustering method on a real case study of spatiotemporal EEG data, recorded from three groups of subjects during a cognitive task. The clusters were referred back to the brain data for a better understanding of the data and the processes that generated it. The cluster analysis allowed to discover and understand differences on temporal sequences and spatial involvement of brain regions in response to a cognitive task.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Abbott LF (1999) Lapicque’s introduction of the integrate-and-fire model neuron (1907). Brain Res Bull 50:303–304
Bezdek JC, Ehrlich R, Full W (1984) FCM: the fuzzy c-means clustering algorithm. Comput Geosci 10:191–203
Capecci E, Kasabov N, Wang GY (2015) Analysis of connectivity in NeuCube spiking neural network models trained on EEG data for the understanding of functional changes in the brain: a case study on opiate dependence treatment. Neural Netw 68:62–77
Carpenter GA, Grossberg S (1987) A massively parallel architecture for a self-organizing neural pattern recognition machine. Comput Vis Gr Image Process 37:54–115
Carter R (2014) The human brain book. Penguin books
Deboeck G, Kohonen T (1998) Visual explorations in finance: with self-organizing maps. Springer
Delbruck T (2007) jAER open source project. http://sourceforge.net/p/jaer/wiki/Home/. Accessed 15 May 2016
Deng D, Kasabov N (2000) ESOM: an algorithm to evolve self-organizing maps from on-line data streams. IJCNN 6:3–8
Deng M, Liu Q, Wang J, Shi Y (2013) A general method of spatio-temporal clustering analysis. Sci China Inf Sci 56:1–14
Ester M, Kriegel H-P, Sander J, Xu X (1996) A density-based algorithm for discovering clusters in large spatial databases with noise. In: Kdd 96(34):226–231
Gholami Doborjeh M, Kasabov N (2015) Dynamic 3D clustering of spatio-temporal brain data in the NeuCube spiking neural network architecture on a case study of fMRI data. In: ICONIP, pp 191–198
Gholami Doborjeh M, Kasabov N (2016) Personalised modelling on integrated clinical and EEG spatio-temporal brain data in the NeuCube spiking neural network system. In: WCCI: IJCNN, pp 1373–1378
Gholami Doborjeh M, Wang GY, Kasabov N, Kydd R, Russell B (2016) A spiking neural network methodology and system for learning and comparative analysis of EEG data from healthy versus addiction treated versus addiction not treated subjects. IEEE Trans Biomed Eng 63(9):1830–1841
Hartigan JA, Wong M (1979) Algorithm AS 136: a k-means clustering algorithm. J R Stat Soc Ser C Appl Stat 28:100–108
Hassinen P (2015) SOM analyzer. http://som-analyzer.software.informer.com/. Accessed 15 May 2016
Hebb DO (1949) The organization of behavior: a neuropsychological approach. Wiley
Indiveri G et al (2011) Neuromorphic silicon neuron circuits. Front Neurosci 5:1–23, Article 73
Johnson SC (1967) Hierarchical clustering schemes. Psychometrika 32:241–254
Kasabov N (2001) Evolving fuzzy neural networks for supervised/unsupervised online knowledge-based learning. IEEE Trans Syst Man Cybern Part B Cybern 31:902–918
Kasabov N (2007) Evolving connectionist systems. Springer
Kasabov N (2012) NeuCube EvoSpike architecture for spatio-temporal. In: Artificial neural networks in pattern recognition, lecture notes in computer science. Springer, pp 225–243
Kasabov N (2014) NeuCube: a spiking neural network architecture for mapping, learning and understanding of spatio-temporal brain data. Neural Netw 52:62–76
Kasabov N, Capecci E (2015) Spiking neural network methodology for modelling, classification and understanding of EEG spatio-temporal data measuring cognitive processes. Inf Sci 294:565–575
Kasabov N, Song Q (2002) DENFIS: dynamic evolving neural-fuzzy inference system and its application for time-series prediction. IEEE Trans Fuzzy Syst 10:144–154
Kasabov N, Dhoble K, Nuntalid N, Indiveri G (2013) Dynamic evolving spiking neural networks for on-line spatio-and spectro-temporal pattern recognition. Neural Netw 41:188–201
Kasabov N, Scott NM, Tu E, Marks S (2016) Evolving spatio-temporal data machines based on the NeuCube neuromorphic framework: design methodology and selected applications. Neural Netw 78:1–14
Katwal SB, Gore JC, Marois R, Rogers BP (2013) Unsupervised spatiotemporal analysis of FMRI data using graph-based visualizations of self-organizing maps. IEEE Trans Biomed Eng 60:2472–2483
Koessler L, Maillard L, Benhadid A, Vignal JP, Felblinger J, Vespignani H, Braun M (2009) Automated cortical projection of EEG sensors: anatomical correlation via the international 10–10 system. Neuroimage 46:64–72
Kohonen T (1998) The self-organizing map. Neurocomputing 21:1–6
Lancaster JL et al (2000) Automated Talairach atlas labels for functional brain. Hum Brain Mapp 10:120–131
Lancaster JL et al (2007) Bias between MNI and Talairach coordinates analyzed using the ICBM-152 brain template. Hum Brain Mapp 28:1194–1205
Liao W, Chen H, Yang Q, Lei X (2008) Analysis of fMRI data using improved self-organizing mapping and spatio-temporal metric hierarchical clustering. IEEE Trans Med Imaging 27:1472–1483
Maass W, Thomas N, Henry M (2002) Real-time computing without stable states: a new framework for neural computation based on perturbations. Neural Comput 14:2531–2560
Mirkin B (1998) Mathematical classification and clustering. Springer
Niedermeyer E, da Silva FL (2005) Electroencephalography: basic principles, clinical applications, and related fields. Lippincott Williams and Wilkins
Ogawa S, Tank DW, Menon R, Ellermann JM, Kim SG, Merkle H, gurbil K (1992) Intrinsic signal changes accompanying sensory stimulation: functional brain mapping with magnetic resonance imaging. Proc Natl Acad Sci 89:5951–5955
Schliebs S, Fiasche M (2012) Constructing robust liquid state machines to process highly variable data streams. In: International Conference on Artificial Neural Networks, Springer, pp 604–611
Schliebs S, Capecci E, Kasabov N (2013) Spiking neural network for on-line cognitive activity classification based on EEG data. In: International Conference on Neural Information Processing, Springer, pp 55–62
Song S, Miller KD, Abbott LF (2000) Competitive Hebbian learning through spike-timing-dependent synaptic plasticity. Nat Neurosci 3:919–926
Talairach J, Tournoux P (1988) Co-planar stereotaxic atlas of the human brain. 3-Dimensional proportional system: an approach to cerebral imaging. Thieme Medical Publishers, New York
Tanay A, Sharan R, Shamir R (2002) Discovering statistically significant biclusters in gene expression data. Bioinformatics 18:136–144
Thorpe S, Gautrais J (1998) Rank order coding. In: Computational neuroscience, Springer, pp 113–118
Tu E, Cao L, Yang J, Kasabov N (2014) A novel graph-based k-means for nonlinear manifold clustering and representative selection. Neurocomputing 143:109–122
Tu E, Kasabov N, Yang J (2016) Mapping temporal variables into the NeuCube for improved pattern recognition, predictive modelling and understanding of stream data. IEEE Trans Neural Netw Learn Syst. doi:10.1109/TNNLS.2016.2536742
Wang W, Yang J, Muntz R (1997) STING: a statistical information grid approach to spatial data mining. VLDB 97:186–195
Zhou D, Bousquet O, Lal TN, Weston J (2004) Learning with local and global consistency. Adv Neural Inf Process Syst 16:321–328
Acknowledgements
The research is supported by the Knowledge Engineering and Discovery Research Institute of the Auckland University of Technology (http://www.kedri.aut.ac.nz). M. GD was also supported by a summer research grant from the faculty of Design and Creative Technology of AUT. The authors would like to acknowledge Professor Robert Kydd and Dr. Bruce Russell from the University of Auckland and Dr. Grace Wang from AUT for providing us with the EEG data. We also acknowledge the assistance of Joyce D’Mello, Dr. Enmei Tu, Dr. Elisa Capecci, Lei Zhou, Israel Espinosa Ramos and Akshay Gollahalli. We are indebted to the reviewers for their detailed, precise and constructive comments and suggestions that helped us tremendously. The NeuCube software along data are available free at http://www.kedri.aut.ac.nz/neucube.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Doborjeh, M.G., Kasabov, N. & Doborjeh, Z.G. Evolving, dynamic clustering of spatio/spectro-temporal data in 3D spiking neural network models and a case study on EEG data. Evolving Systems 9, 195–211 (2018). https://doi.org/10.1007/s12530-017-9178-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12530-017-9178-8