|
|||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | ||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |
java.lang.Objectcreate_NN
public class create_NN
Field Summary | |
---|---|
static double[] |
data_max
pole maximalnych hodnot jednotlivych vstupnych premennych |
static double[] |
data_min
pole minimalnych hodnot jednotlivych vstupnych premennych |
static double |
error
chyba MSE natrenovanej siete |
static double[][] |
input
pole vstupnych premennych |
static int |
num_data
pocet vzoriek v datach nacitanych zo suboru |
static int |
num_inputs
pocet vstupnych premennych |
static double[][] |
output
pole vystupnych premennych |
static int |
size
|
Constructor Summary | |
---|---|
create_NN()
|
Method Summary | |
---|---|
static org.encog.neural.networks.BasicNetwork |
create_RPROP(java.io.File file,
int maxEpoch,
double minError,
int num_Hneuron)
Class Funkcia vytvara novu doprednu siet s uciacim algoritmom resilient propagation |
static double |
denormalize(double number,
double d_min,
double d_max)
Class Denormalizuje hodnotu z intervalu <-1,1> na realnu hodnotu |
static double |
normalize(double number,
double d_min,
double d_max)
Class Normalizuje realnu hodnotu do intervalu <-1,1> |
static void |
read_csv_file(java.io.File file)
Class Nacitava data zo siete a uklada vstupne premenne do pola input a vystupne do pola output |
static void |
recognize_file_outputs(java.io.File file,
java.lang.String file_out,
org.encog.neural.networks.BasicNetwork network1)
Class Funkcia rata vystupy pre sady zadanych vstupnych premennych ulozenych v subore cvs a zapisuje vstupy spolu s vystupmi do suboru cvs do pola input a vystupne do pola output |
static double |
recognize_output(double[] inputs,
org.encog.neural.networks.BasicNetwork network1)
Class Funkcia rata vystup pri zadanych vstupnych premennych a natrenovanej sieti |
static double |
round(double x,
int pocet)
Class Zaokruhluje hodnotu na dany pocet desatinych miest |
Methods inherited from class java.lang.Object |
---|
equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
Field Detail |
---|
public static int num_data
public static int num_inputs
public static int size
public static double error
public static double[] data_max
public static double[] data_min
public static double[][] input
public static double[][] output
Constructor Detail |
---|
public create_NN()
Method Detail |
---|
public static double normalize(double number, double d_min, double d_max)
double
- Konkretne cislo na normalizaciudouble
- Minimalna hodnota intervalu, z ktoreho je cislodouble
- Maximalna hodnota intervalu, z ktoreho je cislo
public static double denormalize(double number, double d_min, double d_max)
double
- Konkretne cislo na denormalizaciudouble
- Minimalna hodnota intervalu, do ktoreho chceme denormalizovatdouble
- Maximalna hodnota intervalu, do ktoreho chceme denormalizovat
public static void read_csv_file(java.io.File file) throws java.io.IOException
File
- Subor, v ktorom mame ulozene data na trenovanie siete
java.io.IOException
public static org.encog.neural.networks.BasicNetwork create_RPROP(java.io.File file, int maxEpoch, double minError, int num_Hneuron) throws java.io.IOException
File
- Subor, v ktorom mame ulozene trenovacie dataint
- Maximalny pocet epochov pri trenovani sietedouble
- Minimalna chcena chyba MSE siete, podmienka ukoncenia trenovania sieteint
- Pocet neuronov v skrytej vrstve, minimalne dvojnasobny pocet
vstupnych premennych( ak je pocet mensi, defaulte sa nastavi na dvojnasobok)
java.io.IOException
- Chyba pri citani pri suborepublic static double round(double x, int pocet)
double
- Cislo na zaokruhlovanieint
- Pocet desatinych miestpublic static double recognize_output(double[] inputs, org.encog.neural.networks.BasicNetwork network1)
double[][]
- Pole vstupnych premennychBasicNetwork
- Natrenovana siet, ktora bude ratat vystup
public static void recognize_file_outputs(java.io.File file, java.lang.String file_out, org.encog.neural.networks.BasicNetwork network1) throws java.io.IOException
File
- Subor, v ktorom mame ulozene data, sady vstupnych premennych
na ratanie outputu (kazdy riadok obsahuje novu sadu premennych)String
- Nazov, pre subor so zapisanymi vyratami vystupnymi hodnotamiBasicNatwork
- Natrenovana siet ratajuca vystupy
java.io.IOException
- Chyba pri citani pri subore
|
|||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | ||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |