Exploiting Patterns for Handling Incomplete Coevolving EEG Time Series
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The electroencephalogram (EEG) time series is a measure of electrical activity received from multiple electrodes placed on the scalp of a human brain. It provides a direct measurement for characterizing the dynamic aspects of brain activities. These EEG signals are formed from a series of spatial and temporal data with multiple dimensions. Missing data could occur due to fault electrodes. These missing data can cause distortion, repudiation, and further, reduce the effectiveness of analyzing algorithms. Current methodologies for EEG analysis require a complete set of EEG data matrix as input. Therefore, an accurate and reliable imputation approach for missing values is necessary to avoid incomplete data sets for analyses and further improve the usage of performance techniques. This research proposes a new method to automatically recover random consecutive missing data from real world EEG data based on Linear Dynamical System. The proposed method aims to capture the optimal patterns based on two main characteristics in the coevolving EEG time series: namely, (i) dynamics via discovering temporal evolving behaviors, and (ii) correlations by identifying the relationships between multiple brain signals. From these exploits, the proposed method successfully identifies a few hidden variables and discovers their dynamics to impute missing values. The proposed method offers a robust and scalable approach with linear computation time over the size of sequences. A comparative study has been performed to assess the effectiveness of the proposed method against interpolation and missing values via Singular Value Decomposition (MSVD). The experimental simulations demonstrate that the proposed method provides better reconstruction performance up to 49% and 67% improvements over MSVD and interpolation approaches, respectively.
Abstract
The electroencephalogram (EEG) time series is a measure of electrical activity received from multiple electrodes placed on the scalp of a human brain. It provides a direct measurement for characterizing the dynamic aspects of brain activities. These EEG signals are formed from a series of spatial and temporal data with multiple dimensions. Missing data could occur due to fault electrodes. These missing data can cause distortion, repudiation, and further, reduce the effectiveness of analyzing algorithms. Current methodologies for EEG analysis require a complete set of EEG data matrix as input. Therefore, an accurate and reliable imputation approach for missing values is necessary to avoid incomplete data sets for analyses and further improve the usage of performance techniques. This research proposes a new method to automatically recover random consecutive missing data from real world EEG data based on Linear Dynamical System. The proposed method aims to capture the optimal patterns based on two main characteristics in the coevolving EEG time series: namely, (i) dynamics via discovering temporal evolving behaviors, and (ii) correlations by identifying the relationships between multiple brain signals. From these exploits, the proposed method successfully identifies a few hidden variables and discovers their dynamics to impute missing values. The proposed method offers a robust and scalable approach with linear computation time over the size of sequences. A comparative study has been performed to assess the effectiveness of the proposed method against interpolation and missing values via Singular Value Decomposition (MSVD). The experimental simulations demonstrate that the proposed method provides better reconstruction performance up to 49% and 67% improvements over MSVD and interpolation approaches, respectively.
참고문헌 (29)
- N. F. Guler, "Recurrent neural networks employing Lyapunov exponents for EEG signals classification", Expert System Appl, 29, pp.506~514 (2005)
- K. Polat, "Classification of epileptiform EEG using a hybrid system based on decision tree classifier and fast Fourier transform", Applied Mathematics and Computation-AMC, 187(2), pp.1017~1026 (2006)
- H. Adeli, "Analysis of EEG records in an epileptic patient using wavelet transform", Journal of Neuroscience methods, 123, pp.69~87 (2003)
- N. Sivasankari, "Automated Epileptic Seizure Detection in EEG signals Using FastICA and Neural Network", Int. J. Advances, Soft computer, Appl, 1(2), pp.91~99 (2009)
- N. F Guler, "Adaptive neuro-fuzzy inference system for classification of EEG signals using wavelet coefficient", Journal of Neuroscience methods, 148(2), pp.113~121 (2005)
- N. Kannathal, "Entropies for detection of epilepsy in EEG", Computer Methods Programs Biomed, 80(3), pp.187~194 (2005)
- A. Subasi, "EEG signal classification using wavelet feature extraction and a mixture of expert model", Expert Systems with Applications, 32(4), pp.1084~1093 (2007)
- E. Acar, "Scalable Tensor Factorizations with Missing Data", Proceedings of the 2010 SIAM International Conference on Data Mining, pp.701~702 (2010)
- A. Sorjamaa, "Methodology for long-term prediction of time series", Neurocomputing, 70, pp.16~18 (2007)
- C. Chatfield, "The Analysis of Time Series: An Introduction" (2008)
- C. F. Gerald, "Applied Numerical Analysis in 7th Edition" (2007)
- G. Liu, "Estimation of missing markers in human motion capture", International Journal of Computer Graphics, 22(9), pp.721~728 (2009)
- S. I. Park, "Capturing and animating skin deformation in human motion", ACM Trans. Graph, 25(3), pp.881~889 (2006)
- M. Kurucz, "Method for large scales SVD with missing values", KDDCup. 07 (2007)
- S. H. Kim, "Incremental expectation maximization principal component analysis for missing value imputation for coevolving EEG data", Journal of Zhejiang University-SCIENCE C (Computer & Electronics), 12(8), pp.687~697 (2011)
- S. Guha, "Correlating synchronous and asynchronous data stream", Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining (2003), KDD
- S. Papadimitriou, "Streaming pattern discovery in multiple time series", Proceeding of the 31st VLDB (2005)
- N. S Nati, "Weighted low-rank approximation", 20th International Conference on Machine Learning, pp.720~727 (2003), AAAI Press
- O. Troyanskaya, "Missing value estimation methods for DNA microarrays", Bioinformatics, 7, pp.520~525 (2001)
- C. Phong, "Missing value estimation for time series microarray data using linear dynamical systems modeling", Advanced Information Networking and Applications-AINA, pp.814~819 (2008)
- A. Aristidou, "Predicting missing marker to drive real- time center of rotation estimation", AMDO ’08: Proceeding of the 5th international conference on articulated motion and deformable objects, pp.238~249 (2008), Springer-Verlarg
- Robust Optical User Motion Tracking Using a Kalman Filter, "K. Dorfmuller-Ulhaas", Technical Report of Augsburg University, ~ (2003)
- R. E. Kalman, "A New Approach to Linear Filtering and Prediction Problems", Journal of Basic Engineering, 81, pp.35~45 (1960)
- L. Ralaivola, "Time series Filtering, Smoothing and Learning using the Kernel Kalman Filter", Proceedings of International Joint Conference on Neural Network (2005)
- L. Li, "Cut- And- Stitch: Efficient Parallel Learning of Linear Dynamical Systems on SMPs", KDD’08 (2008)
- L. Li, "Dynammo: Mining and summarization of coevolving sequences with missing values", KDD (2009)
- R. H. Shumway, "An approach to time series smoothing and forecasting using the EM algorithm", Journal of Time Series Analysis, 3, pp.253~264 (1982)
- R. G. Andrzejak, "Indications of nonlinear deterministic and finite dimensional structures in time series of brain electrical activity", Physical Review E (2001)
- B. Blankertz, "Classifying Single Trial EEG: Towards Brain Computer Interfacing", Advances in Neural Inf (2002)