Abstract | ||
---|---|---|
In this paper, we use a real valued genetic algorithm (GA) to model a large noisy periodic signal. The information that must be extracted are the amplitude, angular velocity and phase of the sines composing the signal. The algorithm outperforms the Fourier Transform (FT) method which has limitations when it comes to determining the phase component of a real part signal. The GA returns the phase component of the signal without the need of de-noising in a very short time, thanks to an efficient parallelization on GPGPU cards (≈ X400 acceleration compared to the same sequential code).
|
Year | DOI | Venue |
---|---|---|
2020 | 10.1145/3377929.3390008 | GECCO '20: Genetic and Evolutionary Computation Conference
Cancún
Mexico
July, 2020 |
DocType | ISBN | Citations |
Conference | 978-1-4503-7127-8 | 0 |
PageRank | References | Authors |
0.34 | 0 | 8 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ulviya Abdulkarimova | 1 | 0 | 0.68 |
Igor Santos Peretta | 2 | 0 | 0.34 |
Anna Ouskova Leonteva | 3 | 0 | 0.68 |
Younes Monjid | 4 | 0 | 0.34 |
Rabih Amhaz | 5 | 0 | 0.34 |
Marc Haegelin | 6 | 0 | 0.34 |
Pierre Collet | 7 | 62 | 11.26 |
Christian Rolando | 8 | 0 | 0.34 |