TY  - CHAP
U1  - Konferenzveröffentlichung
A1  - Hermann, Matthias
A1  - Umlauf, Georg
A1  - Franz, Matthias O.
T1  - Large-scale independent component analysis by speeding up Lie group techniques
T2  - International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2022, 22 - 27 May, Singapore
N2  - We are interested in computing a mini-batch-capable end-to-end algorithm to identify statistically independent components (ICA) in large scale and high-dimensional datasets. Current algorithms typically rely on pre-whitened data and do not integrate the two procedures of whitening and ICA estimation. Our online approach estimates a whitening and a rotation matrix with stochastic gradient descent on centered or uncentered data. We show that this can be done efficiently by combining Batch Karhunen-Löwe-Transformation [1] with Lie group techniques. Our algorithm is recursion-free and can be organized as feed-forward neural network which makes the use of GPU acceleration straight-forward. Because of the very fast convergence of Batch KLT, the gradient descent in the Lie group of orthogonal matrices stabilizes quickly. The optimization is further enhanced by integrating ADAM [2], an improved stochastic gradient descent (SGD) technique from the field of deep learning. We test the scaling capabilities by computing the independent components of the well-known ImageNet challenge (144 GB). Due to its robustness with respect to batch and step size, our approach can be used as a drop-in replacement for standard ICA algorithms where memory is a limiting factor.
KW  - ICA
KW  - Lie group
KW  - ADAM
Y1  - 2022
SN  - 978-1-6654-0540-9
SB  - 978-1-6654-0540-9
SN  - 978-1-6654-0541-6
SB  - 978-1-6654-0541-6
U6  - https://doi.org/10.1109/ICASSP43922.2022.9746444
DO  - https://doi.org/10.1109/ICASSP43922.2022.9746444
N1  - Volltext im Campusnetz via Datenbank IEEE Xplore abrufbar.
SP  - 4388
EP  - 4392
S1  - 5 Seiten
PB  - IEEE
ER  -