Multi-Layer Perceptrons and Symbolic Data

TL;DR

This paper introduces a recoding method enabling Multilayer Perceptrons to handle symbolic data as inputs and outputs, expanding their applicability to complex real-world problems beyond normed vector spaces.

Contribution

A simple, flexible recoding approach that allows MLPs to process symbolic data, broadening their use in practical applications.

Findings

Recoding method effectively handles symbolic data in MLPs.

Method demonstrated successfully on real-world dataset.

Enhances MLP applicability to complex data types.

Abstract

In some real world situations, linear models are not sufficient to represent accurately complex relations between input variables and output variables of a studied system. Multilayer Perceptrons are one of the most successful non-linear regression tool but they are unfortunately restricted to inputs and outputs that belong to a normed vector space. In this chapter, we propose a general recoding method that allows to use symbolic data both as inputs and outputs to Multilayer Perceptrons. The recoding is quite simple to implement and yet provides a flexible framework that allows to deal with almost all practical cases. The proposed method is illustrated on a real world data set.