PATTERN_SOFTMAX_NORM_CVIP
pattern_softmax_norm_cvip() - two csv files are input, training and test sets, and returns new csv files with the feature vectors normalized with softmax scaling normalization.
Contents
SYNTAX
[normalized_file_tt, normalized_file_tr] = pattern_softmax_norm_cvip(file_tt, file_tr, r)
Input Parameters include :
- file_tt - Name of the test set file. A CSV file with a predefined structure.
- file_tr - Name of the training set file. A CSV file with a predefined structure.
- r - The parameter for softmax scaling.
Output Parameters include :
- normalized_file_tt - A string containing the name of the normalized test file.It is the same name as file_tt with the prefix softmax_.
- normalized_file_tr - A string containing the name of the normalized training file.It is the same name as file_tr with the prefix softmax_.
DESCRIPTION
This function gets the training set and the test set as input arguments and normalizes them using softmax normalization method. Then the results are saved to two new CSV files and their names are returned as output. The softmax method needs an extra parameter that should also be given as input r. The parameter r controls the amount of softmax scaling applied to the vectors.
REFERENCE
1. Scott E Umbaugh. DIGITAL IMAGE PROCESSING AND ANALYSIS: Applications with MATLAB and CVIPtools, 3rd Edition.
EXAMPLE
% Test set file file_tt = 'myTestVectors.CSV'; % Training set file file_tr = 'myTrainingVectors.CSV'; % Calling function [normalized_file_tt, normalized_file_tr] = pattern_softmax_norm_cvip(file_tt, file_tr, 3)
normalized_file_tt = softmax_myTestVectors.CSV normalized_file_tr = softmax_myTrainingVectors.CSV
CREDITS
Author: Mehrdad Alvandipour, March 2017
Copyright © 2017-2018 Scott
E Umbaugh
For updates visit CVIP Toolbox Website