ML MRB1: Difference between revisions
No edit summary |
No edit summary |
||
| Line 18: | Line 18: | ||
---- | ---- | ||
[[Category:INCAR]][[Category:Machine Learning]][[Category:Machine Learned Force Fields]] | [[Category:INCAR]][[Category:Machine Learning]][[Category:Machine Learned Force Fields]][[Category: Alpha]] | ||
Revision as of 15:19, 29 February 2020
ML_FF_MRB1_MB = [integer]
Default: ML_FF_MRB1_MB = ML_FF_MRB2_MB/2
Description: This tag sets the number of radial basis sets used to expand the atomic distribution for the radial descriptor within the machine learning force field method.
The value of ML_FF_MRB1_MB depends on the choice of the cut-off radius (ML_FF_RCUT1_MB and the width of the Gaussian functions used in the broadening of the atomic distributions ML_FF_SION1_MB. The error of the basis calculated on a predetermined grid is calculated on the beginning of the calculations (for details see reference [1]).
References
Related Tags and Sections
ML_FF_LMLFF, ML_FF_MRB2_MB, ML_FF_W1_MB, ML_FF_W2_MB