O’ZBEKISTON RESPUBLIKASI RAQAMLI
TEXNOLOGIYALAR VAZIRLIGI
MUHAMMAD AL-XORAZMIY NOMIDAGI TOSHKENT AXBOROT TEXNOLOGIYALARI UNIVERSITETI
5-AMALIY ISH
MASHINALI O’QITISHGA KIRISH
KIBERXAVFSIZLIK FAKULTETI
712-21 guruh talabasi
Bajardi: Aliqulov Azizjon
Tekshirdi: Ochilov Mannon
Toshkent 2023
5-AMALIY TOPSHIRIQ
Mavzu: Sun’iy neyron tarmoqlari. Sodda neyron tarmoqlarini qurish. Neyron tarmoqlarining to‘gri va teskari tarqalish algoritmlari
3
|
Uy hayvonlarini sinflashtirish
|
60
|
3
|
4
|
1. Ishni bajarish uchun kerakli kutubxonalarni chaqirib olamiz
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
df = pd.read_csv('pet2.csv')
df.head()
Datasetimizni dastlabki 5 elementini chiqaramiz
x = df.iloc[:, [0,1]].values
y = df.iloc[:, 1].values
c=df.iloc[:,-1].values
Yaratilgan dataset ning ixtiyoriy ikkita xususiyatini (Bo’yi vaVazni ) olgan holda matplotlib kutubxonasidan foydalanib grafik tasviri.
plt.xlabel(‘Yoshi’)
plt.ylabel('Vazni')
plt.scatter(x[:,0], x[:,1], marker='^',color='green')
plt.legend()
plt.show()
from keras.utils import to_categorical
y=to_categorical(Y).astype(int)
Yaratilgan datasetni modelni o’qitish uchun 90 % va testlash uchun 10% nisbatda bo’laklarga ajratamiz
from sklearn.model_selection import train_test_split
train_x, test_x, train_y, test_y = train_test_split(X, y, test_size=0.1)
train_x.shape
test_y.shape
Keras kutubxonasidan foydalanib masalaga mos neyron tarmoq arxitekturasini quramiz
from keras.models import Sequential
from keras.layers import Dense
from keras.optimizers import Adam,SGD
model = Sequential()
model.add(Dense(8, input_dim=4, activation='relu', name='fc1'))
model.add(Dense(10, activation='relu', name='fc2'))
model.add(Dense(4, activation='softmax', name='output'))
print('Neyron tarmoq aritekturasi:')
print(model.summary())
Neyon tarmoqni o’qitish paramertlarini(o’qish qadami-lr, o’qitishlar soni-epoch) tanlash
optimizer = Adam(learning_rate=0.001)
model.compile(optimizer, loss='categorical_crossentropy', metrics=['accuracy'])
history = model.fit(train_x, train_y, verbose=1, batch_size=2, epochs=100)
Neyron tarmoqning o’qitish natijalarini garfik tarviri
plt.figure(figsize=(12,4))
plt.subplot(121)
plt.plot(history.history['accuracy'])
plt.title('Model accuracy')
plt.ylabel('Accuracy')
plt.xlabel('Epoch')
plt.grid()
plt.subplot(122)
plt.plot(history.history['loss'])
plt.title('Model loss')
plt.ylabel('Loss')
plt.xlabel('Epoch')
plt.grid()
plt.show()
Model aniqligini(o’rgatuvchi tanalama uchun)
results = model.evaluate(train_x, train_y)
print("O'rgatuvchi tanlama uchun yakuniy loss xatoligi: : {:4f}".format(results[0]))
print("O'rgatuvchi tanlama uchun yakuniy aniqliq(accuracy): {:4f}".format(results[1]))
Modelni test to’plam bilan testlash. Modelini test to’plamdagi aniqligi
results = model.evaluate(test_x, test_y)
print("Test to'lam uchun yakuniy loss xatoligi: : {:4f}".format(results[0]))
print("Test to'plam uchun yakuniy aniqliq(accuracy): {:4f}".format(results[1]))
|
