The Lazy Learning Toolbox For use with MATLAB
Version 1.0


The Lazy Learning Toolbox For use with Matlab consists of four
functions, written in C language, that implement the lazy learning
methods for regression developed at IRIDIA, Universite' Libre de
Bruxelles.

The software is part of a larger IRIDIA project, whose goal is the
implementation of a set of local modeling approaches for data analysis
and regression.



Lazy learning is a memory-based technique that, once a query is
received, extracts a prediction interpolating locally the neighboring
examples of the query which are considered relevant according to a
distance measure.  

This toolbox implements a data-driven method to select on a
query-by-query basis the optimal number of neighbors to be considered
for each prediction.

As an efficient way to identify and validate local models, the recursive
least squares algorithm is adopted.

Furthermore, beside the winner-takes-all strategy for model selection, a
local combination of the most promising models is explored.



Before you can use the Toolbox add the full name of the directory "Lazy"
to the MATLAB path. and then compile, from the MATLAB command line, the
functions conLL.c linLL.c quaLL.c clqLL.c

E.G:
		>> mex -O clqLL.c

Please refer to the Toolbox manual (lazy.ps) for other detail on the
functions. 




The Lazy Learning Toolbox --- Version 1.0:
-------------------------------------------------------------------------
        Copyright (c) 1999 by Mauro Birattari & Gianluca Bontempi
-------------------------------------------------------------------------

            Mauro Birattari                   Gianluca Bontempi
                IRIDIA                             IRIDIA 
    Universite' Libre de Bruxelles     Universite' Libre de Bruxelles
           mbiro@ulb.ac.be                    gbonte@ulb.ac.be

-------------------------------------------------------------------------