How nikhileshnarkhede/symboptai works
SymbOptAI is a research workflow for discovering and using interpretable equations to predict and optimize mechanical properties of 3D-printed materials. Its documented path runs from experimental design and bootstrap augmentation through symbolic regression, equation analysis, feature-enhanced neural-network training, and NSGA-II inverse design. Equation verification and feature-importance analysis are additional analysis capabilities. The README establishes the overall stage sequence; sampled source confirms formula parsing and feature generation within the DNN pipeline. Unsampled modules and integrations are not assumed to call one another.
Main entry points: final_FE_DNN_pipeline.py Feature-enhanced DNN · Verification_of_equations.py Equation verification · Equations_Generator.py Symbolic regression · 4.1 Eqation _verification and Terms analysis Equation analysis
Components 15
Data preparation
- Taguchi DOE
1. DOE/doe.py - Bootstrap validation
2. Synthetic Data Generation and Validation/synthetic_data_generation_and_validation.py
Symbolic equations
- Symbolic regression
3. SRA Equation Grenrator/Equations_Generator.py - Equation analysis
4.1 Eqation _verification and Terms analysis - Equation term parser
4.1 Eqation _verification and Terms analysis/Equation_Terms.py - Term normalization tools
4.1 Eqation _verification and Terms analysis/canonicalize_term_full_Recursive_normalization_with_Fun_formula.py - Term calculation
5. DNN with Feature Engineering/Term_wise_calculation.py - Equation verification
4.1 Eqation _verification and Terms analysis/Verification_of_equations.py
Prediction analysis
- Feature importance
4.2 Feature Importances for Each Property/feature_importances_for_each_property.py - Feature-enhanced DNN
5. DNN with Feature Engineering/final_FE_DNN_pipeline.py - DNN model utilities
5. DNN with Feature Engineering/model_FE_DNN.py
Inverse design
- NSGA-II optimization
6. NSGA 2 optimization/Optimization.py
Other components
- Researcher
- Six property predictions
- Optimized settings
Connections 18
- Researcher → Taguchi DOE: specifies factors
README.md - Taguchi DOE → Bootstrap validation: designs experiments
README.md - Bootstrap validation → Symbolic regression: augments data
README.md - Symbolic regression → Equation analysis: produces equations
README.md - Equation analysis → Feature-enhanced DNN: engineers features
README.md - Feature-enhanced DNN → NSGA-II optimization: supplies predictions
README.md - NSGA-II optimization → Optimized settings: identifies settings
README.md - Researcher → Equation verification: validates equations
README.md - Symbolic regression → Equation verification: equations verified
README.md - Researcher → Feature importance: analyzes features
README.md - Feature importance → Six property predictions: ranks importance
- Feature-enhanced DNN → Six property predictions: predicts properties
README.md - Feature-enhanced DNN → Equation term parser: parses equations
5. DNN with Feature Engineering/final_FE_DNN_pipeline.py - Feature-enhanced DNN → Term calculation: builds term features
5. DNN with Feature Engineering/final_FE_DNN_pipeline.py - Equation verification → Equation term parser: analyzes equations
4.1 Eqation _verification and Terms analysis/Verification_of_equations.py - Equation verification → Term calculation: evaluates terms
4.1 Eqation _verification and Terms analysis/Verification_of_equations.py - Equation analysis → Term normalization tools: normalizes terms
4.1 Eqation _verification and Terms analysis/Terms_counter_full_Recursive_normalization_with_Fun_formula.py - Feature-enhanced DNN → DNN model utilities: model utilities