II.
LibraryProcess overview
Reference · livelib-process:materials-science--ml-property-prediction
ml-property-prediction overview
ML Materials Property Prediction - Build and deploy machine learning models for property prediction using descriptors, graph neural networks (CGCNN, MEGNet), or SISSO.
Attributes
displayName
ml-property-prediction
description
ML Materials Property Prediction - Build and deploy machine learning models for property prediction
using descriptors, graph neural networks (CGCNN, MEGNet), or SISSO.
libraryPath
library/specializations/domains/science/materials-science/ml-property-prediction.js
specialization
materials-science
references
- - matminer: https://hackingmaterials.lbl.gov/matminer/ - CGCNN: https://github.com/txie-93/cgcnn - MEGNet: https://github.com/materialsvirtuallab/megnet - SISSO: https://github.com/rouyang2017/SISSO
example
const result = await orchestrate('domains/science/materials-science/ml-property-prediction', {
targetProperty: 'band-gap',
trainingData: { source: 'materials-project', count: 10000 },
modelType: 'cgcnn'
});
usesAgents
- data-collector
- feature-engineer
- data-splitter
- ml-trainer
- model-evaluator
- importance-analyst
- model-interpreter
- model-exporter
- technical-writer
Outgoing edges
lib_applies_to_domain1
- domain:materials-science·DomainMaterials Science
lib_belongs_to_specialization1
- specialization:materials-science·SpecializationMaterials Science
lib_implements_workflow1
- workflow:experiment-design·WorkflowExperiment Design
lib_involves_role1
- role:research-engineer·RoleResearch Engineer
lib_requires_skill_area3
- skill-area:data-analysis·SkillAreaData Analysis
- skill-area:statistical-analysis·SkillAreaStatistical Analysis
- skill-area:mathematical-reasoning·SkillAreaMathematical Reasoning
uses_agent3
- lib-agent:data-science-ml--feature-engineer·LibraryAgentfeature-engineer
- lib-agent:data-science-ml--model-evaluator·LibraryAgentmodel-evaluator
- lib-agent:meta--technical-writer·LibraryAgenttechnical-writer
Incoming edges
None.