Requests for technical support from the VASP group should be posted in the VASP-forum.

Difference between revisions of "Category:Machine Learning"

From Vaspwiki
Jump to navigationJump to search
 
(11 intermediate revisions by the same user not shown)
Line 1: Line 1:
This section describes how to use machine learning to generate force fields. First, one should become familiar with the theoretical background {{TAG|On-the-fly machine learning force field generation using Bayesian linear regression}}. Secondly, one can focus on how to set up and run the machine learning calculations within VASP {{TAG|Machine learning force field calculations: Basics}}. Finally, one gains some hands-on experience with the following tutorial: {{TAG|Liquid Si - MLFF}}.
+
This section describes the methodology used for force-field generation using machine learning. One first should checkout the theoretical background {{TAG|Machine learning force field: Theory}} together with the basic description how to run the machine learning calculations {{TAG|Machine learning force field calculations: Basics}}. Then gain some hands-on experience with the following tutorial: {{TAG|Liquid Si - MLFF}}.
  
 
== Input ==
 
== Input ==
  
 
*Besides the usual input files ({{TAG|INCAR}}, {{TAG|POSCAR}}, etc.) the machine learning force field method requires the following input files:
 
*Besides the usual input files ({{TAG|INCAR}}, {{TAG|POSCAR}}, etc.) the machine learning force field method requires the following input files:
**{{TAG|ML_ABCAR}} Ab initio data used to create the training data.
+
**{{TAG|ML_AB}} Ab initio data used to create the training data.
**{{TAG|ML_FFCAR}} File containing force field parameters.
+
**{{TAG|ML_FF}} File containing force field parameters.
  
 
== Output ==
 
== Output ==
Line 11: Line 11:
 
*The machine learning force field method generates the following output files:
 
*The machine learning force field method generates the following output files:
 
**{{TAG|ML_LOGFILE}} Main output file for the machine learning force field method.
 
**{{TAG|ML_LOGFILE}} Main output file for the machine learning force field method.
**{{TAG|ML_ABNCAR}} New abinitio data (used as {{TAG|ABCAR}} in the next run).
+
**{{TAG|ML_ABN}} New abinitio data (used as {{TAG|ML_AB}} in the next run).
**{{TAG|ML_REGCAR}} Output file summarizing regression results.
+
**{{TAG|ML_REG}} Output file summarizing regression results.
**{{TAG|ML_HISCAR}} Output file summarizing the histogram data.
+
**{{TAG|ML_HIS}} Output file summarizing the histogram data.
**{{TAG|ML_FFNCAR}} File containing new force field parameters (used as {{TAG|FFCAR}} in the next run).
+
**{{TAG|ML_FFN}} File containing new force field parameters (used as {{TAG|ML_FF}} in the next run).
 
**{{TAG|ML_EATOM}} Output file containing local atomic energies
 
**{{TAG|ML_EATOM}} Output file containing local atomic energies
 
**{{TAG|ML_HEAT}} Output file including local heat flux
 
**{{TAG|ML_HEAT}} Output file including local heat flux
 +
  
  
Line 22: Line 23:
  
 
== Theoretical Background ==
 
== Theoretical Background ==
*{{TAG|On-the-fly machine learning force field generation using Bayesian linear regression}}.
+
*{{TAG|Machine learning force field: Theory}}.
  
 
== How to ==
 
== How to ==
Line 29: Line 30:
  
 
== Tutorial ==
 
== Tutorial ==
*Basic tutorial on how to perform on-the-fly learning and how to control the accuracy of the force field: {{TAG|Liquid Si - MLFF}}.
+
*Basic tutorial to learn how to perform on-the-fly learning and how to control the accuracy of the force field: {{TAG|Liquid Si - MLFF}}.
  
 
----
 
----
[[Category:VASP|Machine Learning]][[Category:VASP6]]
+
[[Category:VASP|Machine Learning]][[Category:Alpha]]

Latest revision as of 15:02, 1 September 2020

This section describes the methodology used for force-field generation using machine learning. One first should checkout the theoretical background Machine learning force field: Theory together with the basic description how to run the machine learning calculations Machine learning force field calculations: Basics. Then gain some hands-on experience with the following tutorial: Liquid Si - MLFF.

Input

  • Besides the usual input files (INCAR, POSCAR, etc.) the machine learning force field method requires the following input files:
    • ML_AB Ab initio data used to create the training data.
    • ML_FF File containing force field parameters.

Output

  • The machine learning force field method generates the following output files:
    • ML_LOGFILE Main output file for the machine learning force field method.
    • ML_ABN New abinitio data (used as ML_AB in the next run).
    • ML_REG Output file summarizing regression results.
    • ML_HIS Output file summarizing the histogram data.
    • ML_FFN File containing new force field parameters (used as ML_FF in the next run).
    • ML_EATOM Output file containing local atomic energies
    • ML_HEAT Output file including local heat flux


All INCAR tags belonging to the machine learning force field method start with the prefix ML_FF_. Input tags that are related to the many-body term end with _MB.

Theoretical Background

How to

Tutorial

  • Basic tutorial to learn how to perform on-the-fly learning and how to control the accuracy of the force field: Liquid Si - MLFF.