FLPD is a fuzzy prototype-based machine learning system that facilitates automated learning. It can be used for various applications, including recognition and classification, and helps in the development of intelligent systems.
The induction kernel of the FLPD software is based on the HPI algorithm (Hierarchical Prototype Induction). However, advanced users have the option to design and use their induction algorithm based on fuzzy prototypes as well. FLPD is structured in layers, which contain multiple utilities working at different abstraction levels. This configuration offers users the flexibility, power, and simplicity desired.
The FLPD software is based on a C++ library for fuzzy logic and machine learning, enabling users to represent and manipulate fuzzy information and apply machine learning techniques. Furthermore, the tool offers some key features that make it stand out than the other software available in the market.
The FLPD software includes the necessary classes for representing and manipulating discrete fuzzy sets and, to some extent, continuous fuzzy sets. Specific classes for handling fuzzy sets transformable in LPD (Least Prejudiced Distribution) probability distributions are also included, with the appropriate mechanics for handling the needed transformations automatically. Additionally, the library implements those operations derived from the Mass Assignment Theory, such as fusion of fuzzy sets and conditional probability.
A basic implementation of prototypes is provided, as containers of fuzzy attributes. The operations on fuzzy prototypes derived from the Mass Assignment Theory is also incorporated, i.e., fusion, support, and average normalized support. The tool offers some useful tools for data fuzzification, linguistic covering algorithms, and prototype induction algorithms. The library includes an implementation of the Hierarchical Prototype Induction algorithm (HPI), HPIW (HPI with Weighting).
The software has a basic implementation of data I/O. Users can read and write data in several formats (text, CSV, and XML) with ease. Besides the library, FLPD also includes some front-ends for the command line and a simple graphic interface written in PHP. These interfaces utilize the library to implement some common automatic learning tasks, such as supervised and unsupervised learning.
FLPD offers a range of learning front-ends implemented over a group of back-ends, each performing a specific task. These back-ends calculate one step in the several sequential tasks needed for a learning process to be completed. Using these back-ends directly, a personalized learning process can be designed, completely new or partially based on the calculation infrastructure already provided by the system.
In conclusion, FLPD is an efficient and advanced software solution for automatic learning based on fuzzy prototypes. Its structured layers, various utilities, and flexible calculations make it easy for even inexperienced users to learn and optimize their systems.
Version 3.14: N/A