A machine learning approach for efficient multi-dimensional integration
arXiv:2009.06697 · doi:10.1038/s41598-021-98392-z
Abstract
We propose a novel multi-dimensional integration algorithm using a machine learning (ML) technique. After training a ML regression model to mimic a target integrand, the regression model is used to evaluate an approximation of the integral. Then, the difference between the approximation and the true answer is calculated to correct the bias in the approximation of the integral induced by a ML prediction error. Because of the bias correction, the final estimate of the integral is unbiased and has a statistically correct error estimation. The performance of the proposed algorithm is demonstrated on six different types of integrands at various dimensions and integrand difficulties. The results show that, for the same total number of integrand evaluations, the new algorithm provides integral estimates with more than an order of magnitude smaller uncertainties than those of the VEGAS algorithm in most of the test cases.
17 pages, 1 figure
References in corpus (7)
- Electroweak gauge boson production at hadron colliders through O(alpha_s^2)
- FEWZ 2.0: A code for hadronic Z production at next-to-next-to-leading order
- Adaptive Multidimensional Integration: VEGAS Enhanced
- The W boson production cross section at the LHC through O(alpha_s^2)
- A complete set of in-medium splitting functions to any order in opacity
- Precision modeling of redshift-space distortions from multi-point propagator expansion
- Efficient Monte Carlo Integration Using Boosted Decision Trees and Generative Deep Neural Networks