Using the nextclosure algorithm to extract rules from trained neural networks - application in solar energy systems

Abstract

Due to their capability of dealing with nonlinear problems, Artificial Neural Networks (ANN) are widely used with several purposes. Once trained, they are capable to solve unprecedented situations, keeping tolerable errors in their outputs. However, humans can not assimilate the knowledge kept by those nets, since such knowledge is implicitly represented by their connections weights. So, in order to facilitate the extraction of rules that describe the knowledge of ANN, Formal Concept Analysis (FCA) and the NextClosure algorithm have been used. Such method is presented in this work, combining ANN, FCA and the NextClosure algorithm to compute the minimal implication base (Stem Base). As an example, solar energy systems are the domain application considered here, due to their importance as substitutes of traditional energy systems.

FCANN: An Approach to Knowledge Representation From ANN Through Formal Concept Analysis - Application in the Cold Rolling Process

“…In this section, the steps to extract knowledge from neural nets discussed in [5] and [10] will be presented: 1) Select a process representative data set in order to train the ANN. It is defined as:…”

“…Then, the extraction of all the rules is a difficult task and, consequently, some important rules may be forgotten. In [5] an approach to extract and represent the knowledge of previously trained neural networks through FCA was presented. The main idea of that approach (FCANN) is a join of Formal Concept Analysis [7], [8] and the Next Closure algorithm [9] to extract knowledge from ANN.…”

An Approach to Knowledge Extraction From ANN Through Formal Concept Analysis Computational Tool Proposal: SOPHIANN

“…In [4] and [5], FCA has been used to extract rules from previously trained neural networks. The method presented in [4] consists in a visual analysis of the line diagram in order to extract the rules.…”