Parallel Approach for Time Series Analysis with General Regression Neural Networks

The accuracy on time delay estimation given pairs of irregularly sampled time series is of great relevance in astrophysics. However the computational time is also important because the study of large data sets is needed. Besides introducing a new approach for time delay estimation, this paper presents a parallel approach to obtain a fast algorithm for time delay estimation. The neural network architecture that we use is general Regression Neural Network (GRNN). For the parallel approach, we use Message Passing Interface (MPI) on a beowulf-type cluster and on a Cray supercomputer and we also use the Compute Unified Device Architecture (CUDA™) language on Graphics Processing Units (GPUs). We demonstrate that, with our approach, fast algorithms can be obtained for time delay estimation on large data sets with the same accuracy as state-of-the-art methods.

[1]  Neil D. Lawrence,et al.  Missing Data in Kernel PCA , 2006, ECML.

[2]  P. Murdin,et al.  Encyclopedia of Astronomy and Astrophysics , 2002 .

[3]  R. Casey,et al.  Advances in Pattern Recognition , 1971 .

[4]  Pierre Courrieu,et al.  Fast Computation of Moore-Penrose Inverse Matrices , 2008, ArXiv.

[5]  Keechul Jung,et al.  GPU implementation of neural networks , 2004, Pattern Recognit..

[6]  M. Bartelmann.  Gravitational lensing , 2010, 1010.3829.

[7]  J. Hjorth,et al.  ESTIMATION OF MULTIPLE TIME DELAYS IN COMPLEX GRAVITATIONAL LENS SYSTEMS , 1998 .

[8]  Alessandro Artusi,et al.  Radial Basis Function Networks GPU-Based Implementation , 2008, IEEE Transactions on Neural Networks.

[9]  M. Oguri.  Gravitational Lens Time Delays: A Statistical Assessment of Lens Model Dependences and Implications for the Global Hubble Constant , 2006, astro-ph/0609694.

[10]  W. Daniel Hillis,et al.  Data parallel algorithms , 1986, CACM.

[11]  David E. Goldberg,et al.  Efficient Parallel Genetic Algorithms: Theory and Practice , 2000 .

[12]  Zhongwen Luo,et al.  Artificial neural network computation on graphic process unit , 2005, Proceedings. 2005 IEEE International Joint Conference on Neural Networks, 2005..

[13]  The Q0957+561 Time Delay From Optical Data , 1997 .

[14]  Peter Tiño,et al.  A Kernel-Based Approach to Estimating Phase Shifts Between Irregularly Sampled Time Series: An Application to Gravitational Lenses , 2006, ECML.

[15]  F. A. Seiler,et al.  Numerical Recipes in C: The Art of Scientific Computing , 1989 .

[16]  Markus Harva,et al.  Bayesian Estimation of Time Delays Between Unevenly Sampled Signals , 2008, 2006 16th IEEE Signal Processing Society Workshop on Machine Learning for Signal Processing.

[17]  Gavin Brown,et al.  Diversity in neural network ensembles , 2004 .

[18]  Ronald L. Rivest,et al.  Introduction to Algorithms, Second Edition , 2001 .

[19]  G. Amdhal,et al.  Validity of the single processor approach to achieving large scale computing capabilities , 1967, AFIPS '67 (Spring).

[20]  Pat Langley,et al.  Learning Process Models with Missing Data , 2006, ECML.

[21]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[22]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .

[23]  Donald F. Specht,et al.  A general regression neural network , 1991, IEEE Trans. Neural Networks.

[24]  Microlensing of Lensed Supernovae , 2006, astro-ph/0608391.

[25]  G. Cowan.  Statistical data analysis , 1998 .

[26]  P. Jakobsson,et al.  Microlensing variability in time-delay quasars , 2006, astro-ph/0607133.

[27]  Peter Tiño,et al.  Uncovering delayed patterns in noisy and irregularly sampled time series: An astronomy application , 2009, Pattern Recognit..

[28]  Nello Cristianini,et al.  An Introduction to Support Vector Machines and Other Kernel-based Learning Methods , 2000 .

[29]  Ian T. Nabney,et al.  Netlab: Algorithms for Pattern Recognition , 2002 .

[30]  Norman P. Jouppi,et al.  Readings in computer architecture , 2000 .

[31]  J. Carlos,et al.  Estimating time delays between irregularly sampled time series , 2007 .

[32]  Lakhmi C. Jain,et al.  Radial Basis Function Networks 2: New Advances in Design , 2001 .

[33]  Nello Cristianini,et al.  Kernel Methods for Pattern Analysis , 2004 .

[34]  William Gropp,et al.  Learning from the Success of MPI , 2001, HiPC.

[35]  Peter Tiño,et al.  How accurate are the time delay estimates in gravitational lensing? , 2006, ArXiv.

[36]  R.A. Gonzalez-Grimaldo,et al.  Analysis of Time Series with Artificial Neural Networks , 2008, 2008 Seventh Mexican International Conference on Artificial Intelligence.

[37]  S. Refsdal.  On the possibility of determining Hubble's parameter and the masses of galaxies from the gravitational lens effect , 1964 .

[38]  William H. Press,et al.  The Time Delay of Gravitational Lens 0957+561. I. Methodology and Analysis of Optical Photometric Data , 1992 .

[39]  Y. Lin,et al.  Parallel Computing on a PC Cluster , 2001, ArXiv.

[40]  O. Wucknitz.  Gravitational Lensing , 2007, Large-Scale Peculiar Motions.

[41]  D. Long,et al.  A Robust Determination of the Time Delay in 0957+561A, B and a Measurement of the Global Value of Hubble's Constant , 1996, astro-ph/9610162.