ML LSPARSDES: Difference between revisions
No edit summary |
No edit summary |
||
| Line 1: | Line 1: | ||
{{TAGDEF|ML_FF_LSPARSDES_MB|[logical]|.FALSE.}} | {{TAGDEF|ML_FF_LSPARSDES_MB|[logical]|.FALSE.}} | ||
Description: This tag specifies whether sparsification of the angular descriptors is done within machine learning force fields. | Description: This tag specifies whether sparsification of the angular descriptors is done within machine learning force fields or not. | ||
---- | ---- | ||
Revision as of 07:46, 6 July 2021
ML_FF_LSPARSDES_MB = [logical]
Default: ML_FF_LSPARSDES_MB = .FALSE.
Description: This tag specifies whether sparsification of the angular descriptors is done within machine learning force fields or not.
For details on the theory of angular descriptor sparsification see here.
To use the sparsification of angular descriptors set the following tags:
- ML_FF_LSPARSDES_MB=.TRUE..
- The number of the highest eigenvalues to which the correlation is measured via the leverage scoring: ML_FF_NRANK_SPARSDES_MB.
- The ratio of the selected descriptors to the total number of descriptors: ML_FF_RDES_SPARSDES_MB. This tag controls the extent of the sparsification.
Related Tags and Sections
ML_FF_LMLFF, ML_FF_NRANK_SPARSDES_MB, ML_FF_RDES_SPARSDES_MB