gr_nn1.cgi

Applied Math & Computer Science Lab

Data Analysis, Optimization & Mathematical Modeling, Artificial Intelligence, Neural Net For Everyday Life Applications

AI/Data Mining Links

Online Free Courses

Online Bookstore

Earn Points for $10 Amazon eGift Card

Practical Data Mining Courses Get Certificate of Completion Now for Free

gr_nn1.cgi

Take 4 questions survey to help us better serve our customers.

Choosing Parameters/Model for Stock Data Prediction

Visit online bookstore
View new arrivals on
artificial intelligence, data mining, text mining, applied math

Projects, Source Code,Downloads from this Site
Genetic Agorithms - Genetic Agorithms programming with perl
Bidirectional Associative Memory - type of recurrent neural network. Source code and online demo.
Tic-Tac-Toe Artificial Intelligence (Reinforcement Learning) is applied to Game Programming.
Naive Bayes Classification Document Classification Using Naive Bayes Classifier with Perl

Free online courses
Get certificate of completion
Artificial intelligence, data mining, web mining

################################################
#                                              #
#  General Regression Neural Network   (GRNN)  #
#  Prediction  based on input-target data      #
#                                              #
################################################

$Number_of_training_inputs = 10;

# Dimension of input
$Number_of_points=6;

$Number_of_hidden_units=3;

$sigma=1.2;

for ($i=0; $i<$Number_of_training_inputs; $i++)
{ 
  for ($j=0; $j<$Number_of_points; $j++)
  { 
             
      $data[$i][$j]=0.5*$j ;
  }

$r=rand(1);
     $r=(0.5-$r);

$target[$i] = 0.5*$Number_of_points  + $r;   
}

for ($i=0; $i<$Number_of_points; $i++)
{ 
     $new_data[0][$i]=0.5*($i) ;
}

$S1=0;
$S2=0;

for ($k=0; $k < $Number_of_training_inputs; $k++)
{

for ($zz=0; $zz < $Number_of_hidden_units; $zz++)
  {

$d=0;   
     for ($i=0; $i<$Number_of_points; $i++)
     {

$d=$d+($data[$k][$i]-$new_data[0][$i])**2;

$d=((-1)*$d)*$sigma ;
         $d=exp($d);

} # for each point
$S1=$S1+$d;
$S2=$S2+$d*$target[$k];   
   
 } #z hidden unit

} #k training unit
print "S2= $S2 \n";
print "S1= $S1 \n";

$Out=$S2/$S1;

print "OUT=:   $Out\n\n";