AVERPOIN is an application software which evaluates analytically the representation quality of a set of neural-network inputs. This quality refers to the network ability to generate any output value from any input of a specified set using only positive weights and positive inputs. Each network input of the set is represented by an array (vector) of positive values. AVEPOIN considers that the network calculates each output through the weighted sum of the corresponding input vector values. Hence AVERPOIN implements an algorithm which receives the set of input vectors encoded by an m-by-n matrix C. m denotes the number of input vectors and n is the number of input neurons, thus each row of C represents an input vector. The output of AVERPOIN is a number between 0 and 1 which encodes the suitability (fitness) of C. A fitness of 0 denotes the worst possible representation, and 1 a representation for which the network is able to generate any output value for any input vector. AVERPOIN is also able to represent graphically the set of inputs so that its suitability can by visually assessed.
AVERPOIN consists in a set of source code files written in MATLAB language.
The fitness value returned by AVERPOIN is conceptually comparable to the MATLAB expression rank(C)/size(C,1) but for the case in which the linear combination of C columns are performed only with positive coefficients.
AVERPOIN can be executed by MATLAB and Octave. It has been tested with MATLAB R2012a and Octave 4.0.0. Considering these versions MATLAB provides a significantly faster execution and better graphical representation. The main function for evaluating analytically a cone C, Ir(C), is:
result=int_res(C,plot_flag).
The testing function for evaluating numerically a cone C, Ir_num(C), is:
result=int_res_num(C,n_pts_dim,plot_flag).
See Documentation section for the function parameter description.
The documentation available as of the date of this release is included in the .m files. To obtain the usage information of any function use the "help" command of MATLAB or Octave, especially:
help int_res
or
help int_res_num
The calculation performed by this software is explained and illustrated in the publication https://doi.org/10.3389/fnins.2018.00913:
Carrillo, R. R., Naveros, F., Ros, E., & Luque, N. R. (2018). A Metric for Evaluating Neural Input Representation in Supervised Learning Networks. Frontiers in Neuroscience, 12, 913.
Copyright (C) 2017 Richard R. Carrillo (University of Granada)
AVERPOIN is free software: you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
AVERPOIN is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.
You should have received a copy of the GNU General Public License along with this program (see the files called COPYING.txt and COPYING_LESSER.txt). If not, see http://www.gnu.org/licenses/.
- Richard R. Carrillo: rcarrillo .AT. ugr.es