Non-linear Least Mean Squares Prediction Based on Non-Gaussian Mixtures

Independent Component Analyzers Mixture Models (ICAMM) are versatile and general models for a large variety of probability density functions. In this paper, we assume ICAMM to derive a closed-form solution to the optimal Least Mean Squared Error predictor, which we have named E-ICAMM. The new predictor is compared with four classical alternatives (Kriging, Wiener, Matrix Completion, and Splines) which are representative of the large amount of existing approaches. The prediction performance of the considered methods was estimated using four performance indicators on simulated and real data. The experiment on real data consisted in the recovering of missing seismic traces in a real seismology survey. E-ICAMM outperformed the other methods in all cases, displaying the potential of the derived predictor.

[1]  Addisson Salazar,et al.  Probabilistic Distance for Mixtures of Independent Component Analyzers , 2018, IEEE Transactions on Neural Networks and Learning Systems.

[2]  Xin Li,et al.  Patch-Based Image Processing: From Dictionary Learning to Structural Clustering , 2017 .

[3]  Terrence J. Sejnowski,et al.  ICA Mixture Models for Unsupervised Classification of Non-Gaussian Classes and Automatic Context Switching in Blind Signal Separation , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[4]  Hua-wei Zhou,et al.  Practical Seismic Data Analysis , 2014 .

[5]  Yuedong Wang,et al.  Smoothing Splines: Methods and Applications , 2011 .

[6]  Luis Vergara,et al.  Simple approach to nonlinear prediction , 2001 .

[7]  C. Jutten,et al.  SRF: Matrix completion based on smoothed rank function , 2011, 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[8]  Emmanuel J. Candès,et al.  Exact Matrix Completion via Convex Optimization , 2009, Found. Comput. Math..

[9]  G. Bellefleur,et al.  Feasibility of virtual source reflection seismology using interferometry for mineral exploration: A test study in the Lalor Lake volcanogenic massive sulphide mining area, Manitoba, Canada , 2015 .

[10]  Roger Woodard,et al.  Interpolation of Spatial Data: Some Theory for Kriging , 1999, Technometrics.

[11]  Eero P. Simoncelli,et al.  Image quality assessment: from error visibility to structural similarity , 2004, IEEE Transactions on Image Processing.

[12]  Addisson Salazar.  On statistical pattern recognition in independent component analysis mixture modelling , 2012 .

[13]  Pierre Comon,et al.  Handbook of Blind Source Separation: Independent Component Analysis and Applications , 2010 .

[14]  Ray Fisher,et al.  Applied Seismology: A Comprehensive Guide to Seismic Theory and Application , 2004 .

[15]  Addisson Salazar,et al.  Semi-blind source extraction of atrial activity by combining statistical and spectral features , 2011, Digit. Signal Process..

[16]  Addisson Salazar,et al.  A general procedure for learning mixtures of independent component analyzers , 2010, Pattern Recognit..

[17]  G. A. Einicke,et al.  Smoothing, Filtering and Prediction - Estimating The Past, Present and Future , 2012 .

[18]  N. Wiener,et al.  Nonlinear Problems in Random Theory , 1964 .

[19]  John Etgen,et al.  Strike shooting, dip shooting, widepatch shooting — Does prestack depth migration care? A model study , 1998 .