A modular, object-oriented Java library implementing Neural Networks, Fuzzy Logic, and Genetic Algorithms, supported by real-world case studies demonstrating practical soft computing applications.
NFG-Library provides unified, reusable, and maintainable implementations of core soft computing techniques.
The goal is to offer a clean API that allows developers, students, and researchers to integrate soft computing methods into different applications easily.
This repository also includes case studies (one per phase) that showcase how each technique can be applied to solve real-world problems.
- Rank / Roulette selection
- Crossover and mutation operators
- Fitness evaluation interface
- Customizable parameters
- Fuzzy sets and membership functions
- Fuzzy rules and inference engine
- Defuzzification methods
- Modular rule-based design
- Feed-forward neural networks
- Backpropagation training
- Activation functions (ReLU, Sigmoid, etc.)
- Configurable architectures
- Job Scheduling
- Automatic Window Blind Control
- Object-Oriented Design
- Modular Architecture
- Soft Computing Concepts
