Robust large-scale statistical inference and ICA using bootstrapping

 |  Login

Show simple item record

dc.contributor Aalto-yliopisto fi
dc.contributor Aalto University en
dc.contributor.advisor Ollila, Esa, Prof., Aalto University, Department of Signal Processing and Acoustics
dc.contributor.advisor Finland
dc.contributor.advisor Koivunen, Visa, Prof., Aalto University, Department of Signal Processing and Acoustics
dc.contributor.advisor Finland
dc.contributor.author Basiri, Shahab
dc.date.accessioned 2018-12-05T10:03:52Z
dc.date.available 2018-12-05T10:03:52Z
dc.date.issued 2018
dc.identifier.isbn 978-952-60-8023-9 (printed)
dc.identifier.issn 1799-4934 (printed)
dc.identifier.uri https://aaltodoc.aalto.fi/handle/123456789/34872
dc.description.abstract The reliability of the information extracted from large-scale data, as well as the validity of data-driven decisions depend on the veracity of the data and the utilized data processing methods. Quantification of the veracity of parameter estimates or data-driven decisions is required in order to make appropriate choices of estimators and identifying redundant or irrelevant variables in multi-variate data settings. Moreover, quantification of the veracity allows efficient usage of available resources by processing only as much data as is needed to achieve a desired level of accuracy or confidence. Statistical inference such as finding the accuracy of certain parameter estimates and testing hypotheses on model parameters can be used to quantify the veracity of large-scale data analytics results. In this thesis, versatile bootstrap procedures are developed for performing statistical inference on large-scale data. First, a computationally efficient and statistically robust bootstrap procedure is proposed, which is scalable to smaller distinct subsets of data. Hence, the proposed method is compatible with distributed storage systems and parallel computing architectures. The statistical convergence and robustness properties of the method are analytically established. Then, two specific low-complexity bootstrap procedures are proposed for performing statistical inference on the mixing coefficients of the Independent Component Analysis (ICA) model. Such statistical inferences are required to identify the contribution of a specific source signal-of-interest onto the observed mixture variables. This thesis establishes significant analytical results on the structure of the FastICA estimator, which enable the computation of bootstrap replicas in closed-form. This not only saves computational resources, but also avoids convergence problems, permutation and sign ambiguities of the FastICA algorithm. The developed methods enable statistical inference in a variety of applications in which ICA is commonly applied, e.g., fMRI and EEG signal processing. In the thesis, an alternative derivation of the fixed-point FastICA algorithm is established. The derivation provides a better understanding of how the FastICA algorithm is derived from the exact Newton-Raphson (NR) algorithm. In the original derivation, FastICA was derived as an approximate NR algorithm using unjustified assumptions, which are not required in the alternative derivation presented in this thesis. It is well known that the fixed-point FastICA algorithm has severe convergence problems when the dimensionality of the data and the sample size are of the same order. To mitigate this problem, a power iteration algorithm for FastICA is proposed, which is remarkably more stable than the fixed-point FastICA algorithm. The proposed PowerICA algorithm can be run in parallel on two computing nodes making it considerably faster to compute. en
dc.format.extent 72 + app. 50
dc.language.iso en en
dc.publisher Aalto University en
dc.publisher Aalto-yliopisto fi
dc.relation.ispartofseries Aalto University publication series DOCTORAL DISSERTATIONS en
dc.relation.ispartofseries 107/2018
dc.relation.haspart [Publication 1]: S. Basiri, E. Ollila, V. Koivunen. Fast and robust bootstrap in analyzing large multivariate datasets. In 48th Asilomar Conference on Signals, Systems and Computers, Pacific Grove, USA, pp. 8-13, November 2014. DOI: 10.1109/ACSSC.2014.7094385
dc.relation.haspart [Publication 2]: S. Basiri, E. Ollila, V. Koivunen. Robust, Scalable, and Fast Bootstrap Method for Analyzing Large Scale Data. IEEE Transactions on Signal Processing, vol. 64(4), pp. 1007-1017, February 2015. DOI: 10.1109/TSP.2015.2498121
dc.relation.haspart [Publication 3]: S. Basiri, E. Ollila, V. Koivunen. Alternative derivation of FastICA with novel power iteration algorithm. IEEE Signal Processing Letters, vol. 24(9), pp. 1378-1382, July 2017. DOI: 10.1109/LSP.2017.2732342
dc.relation.haspart [Publication 4]: S. Basiri, E. Ollila, V. Koivunen. Fast and robust bootstrap method for testing hypotheses in the ICA model. In IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Florence, Italy, pp. 6-10, May 2014. DOI: 10.1109/ICASSP.2014.6853547
dc.relation.haspart [Publication 5]: S. Basiri, E. Ollila, V. Koivunen. Enhanced bootstrap method for statistical inference in the ICA model. Signal Processing, vol. 138, 2017, pp. 53-62, March 2017. DOI: 10.1016/j.sigpro.2017.03.005
dc.subject.other Computer science en
dc.title Robust large-scale statistical inference and ICA using bootstrapping en
dc.type G5 Artikkeliväitöskirja fi
dc.contributor.school Sähkötekniikan korkeakoulu fi
dc.contributor.school School of Electrical Engineering en
dc.contributor.department Signaalinkäsittelyn ja akustiikan laitos fi
dc.contributor.department Department of Signal Processing and Acoustics en
dc.subject.keyword Big data analytics en
dc.subject.keyword Bootstrap en
dc.subject.keyword Fast and Robustg Bootstrap en
dc.subject.keyword Distributed and parallel computation en
dc.subject.keyword Robust estimation en
dc.subject.keyword Independent Component Analysis en
dc.subject.keyword FastICA en
dc.identifier.urn URN:NBN:fi:aalto-201812055887
dc.type.dcmitype text en
dc.type.ontasot Doctoral dissertation (article-based) en
dc.type.ontasot Väitöskirja (artikkeli) fi
dc.contributor.supervisor Ollila, Esa, Prof., Aalto University, Department of Signal Processing and Acoustics; Finland
dc.opn Palomar, Daniel, Prof., Hong Kong University of Science and Technology, Hong Kong
dc.opn Tourneret, Jean-Yves, Prof., University of Toulouse, France
dc.rev Palomar, Daniel, Prof., Hong Kong University of Science and Technology, Hong Kong
dc.rev Comon, Pierre, Prof., Université Grenoble Alpes, France
dc.date.defence 2018-06-15


Files in this item

Files Size Format View

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record

Search archive


Advanced Search

article-iconSubmit a publication

Browse

My Account