Journal
2023 27TH INTERNATIONAL CONFERENCE ON METHODS AND MODELS IN AUTOMATION AND ROBOTICS, MMAR
Volume -, Issue -, Pages 240-245Publisher
IEEE
DOI: 10.1109/MMAR58394.2023.10242501
Keywords
Gaussian Process; Chebyshev Polynomials; Chebyshev Series; computing efficiency
Categories
Ask authors/readers for more resources
Gaussian processes have gained popularity as a way to solve statistics and machine learning problems. However, the computational burden can be impractical, especially with big data. To address this issue, we propose an alternative approximation method using Chebyshev polynomials. Our method calculates the GP model only at Chebyshev nodes and transforms function values into Chebyshev coefficients to solve the original problem.
Gaussian processes (GP) are becoming more and more popular way to solve statistics and machine learning problems. One of the reasons is the increase in computational power that can handle the inherent computational problem for GP models. Still, in the case of big data, the computational burden can be impractical. For this reason, various approximation methods are developed. In our work, we would like to present an alternative to the internal approximation by using the properties of Chebyshev polynomials. The idea is to calculate the GP model only at Chebyshev nodes and use the property of transforming function values in them to Chebyshev coefficients giving a solution to the original problem. In our research, we propose our version of the algorithm and test it on cases of various functions.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available