Usage

This is the usage guide for Neuro-Fuzzy Toolbox for building and training neuro-fuzzy models.

If you have not yet installed the package, follow the instructions in Installation.

Important

Neuro-Fuzzy Toolbox is built on top of PyTorch, so a basic familiarity with this library is recommended for a better understanding of its usage (see PyTorch basics).

• Models
  • ANFIS Model Basics
  • ANFIS variants
  • Output Types on ANFIS Models
• Training
  • Initial Considerations
  • Hybrid Learning Algorithm
  • Basic Optimizer Training Algorithm
  • Double Optimizer Training Algorithm
  • SONFIS
  • Custom Training
• Rule Inspection and Analysis
  • 1. Instantiation
  • 2. Inspecting intermediate layer outputs
  • 3. Identifying the most active rules
  • 4. Full prediction explanation
  • 5. Inspecting the global fuzzy structure
  • 6. Sample-specific antecedents