Dhisida qaabka shabkada neerfaha waa meel soo jiidasho leh xagga barashada mashiinka, gaar ahaan Python. Waxay bixisaa baaxad balaadhan oo loogu talagalay falanqaynta, saadaalinta, iyo toosinta hababka go'aan qaadashada. Ka hor inta aynaan u gelin nitty-gritty ee dhisidda shabakad neural ah, waxaa muhiim ah in la fahmo waxa ay tahay shabakad neural ah. Asal ahaan waa nidaam algorithms oo xidhiidhiya qaab-dhismeedka maskaxda bini'aadamka, sidaas darteed abuurista shabakad neerfeed oo macmal ah oo, iyada oo loo marayo habka falanqaynta fasiraadda xogta dareenka, soo qaadashada nuances 'aan la arki karin' xogta ceeriin, si la mid ah maskaxdeenu.
Shabakadda neerfaha ayaa lama huraan u ah geeddi-socodka qodista xogta, halkaas oo ay ku tilmaamto qaababka iyo isbeddellada aad u adag ee aadanaha ama farsamooyinka kale ee kombiyuutarka. Hadda, aan u dhex galno udub dhexaadka arrinta — annagoo adeegsanayna Python si aan u dhisno una sawirno shabakad neerfaha ah.
Isku xirka shabakadaha neural ee Python
# Importing libraries import numpy as np import matplotlib.pyplot as plt from sklearn.datasets import make_blobs # Create a sample dataset dataset=make_blobs(n_samples=800, centers=2, n_features=2, cluster_std=1.6, random_state=50) # Split into input (X) and output (y) X, y = dataset # Plot the sample data plt.scatter(X[:,0], X[:,1], c=y) plt.show()
Aynu fahanno koodkan:
- Afarta sadar ee hore, waxaanu soo dajinaa maktabadaha lagama maarmaanka ah sida numpy, matplotlib iwm.
- Marka xigta, annagoo adeegsanayna shaqada 'make_blobs' ee sklearn, waxaan abuurnaa xog-ururin.
- Kadibna xog-ururinta waxa loo qaybiyaa agab (X) iyo wax-soo-saar (y).
- Sadarka ugu dambeeya wuxuu dhigayaa X iyo y wuxuuna na siinayaa muuqaal xogta iyadoo la adeegsanayo shaqada kala firdhisan ee maktabadda matplotlib.
Fahamka goobta maktabadaha shabakada neural
Fahamka maktabadaha Python ee macnaha guud waa muhim. Maktabadda nambarada waxay ogolaataa hawlgallada xisaabta, matplotlib waxaa loo isticmaalaa 2D garaaf sawiridda xogta ku jirta Python iyo barashada mashiinka warannada sklearn ee Python.
Koodhka talaabo-tallaabo
Habka tallaabo-tallaabo ee koodka ayaa noo ogolaanaya inaan helno faham qoto dheer:
# Import necessary modules from keras.models import Sequential from keras.layers import Dense # Create the model model = Sequential() # Add input layer with 2 inputs neurons model.add(Dense(input_dim=2, output_dim=1, init='uniform', activation='sigmoid')) # Compile model model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy']) # Fit the model history = model.fit(X, y, epochs=100, batch_size=10)
Qaybtan koodka ah,
- Waxaan abuurnaa moodal annagoo adeegsanayna shaqada taxanaha ah () ee ka socota moduleka keras.models.
- Marka xigta, lakabka wax gelinta ayaa lagu daraa 2 neuron oo gashanaya. Halkan, 'Dense' waa nooc lakab ah oo u shaqeeya kiisaska intooda badan. Lakabka cufan, dhammaan qanjidhada lakabka hore waxay ku xidhmaan noodhka lakabka hadda jira.
- 'Compile' waxay diyaarisaa tusaalaha tababarka.
- Qeybta ugu dambeysa, 'ku-habboonaanta moodeelka' waa meesha shabakadda neerfaha lagu tababaro. 'Epochs' waxay tusinaysaa tirada baasaska dhammaan xogta tababarka. Qaabku waxa uu bartaa oo cusboonaysiiyaa halbeegyada moodeelka xilli kasta. Cabbirka dufcaddu waa qayb-hoosaadyada xogta.
Xeerarkan, waxaan ku dhiseynaa aasaaska abuurista shabakad neural ah iyadoo la adeegsanayo Python. Iyada oo la adeegsanayo maktabadaha ballaaran ee Python iyo awoodaha xoogga leh, shabakadaha neerfaha waa la hirgelin karaa oo loo arki karaa si wax ku ool ah. Waxa kaliya oo ku saabsan fahamka xididdada, oo waxaad ku fiican tahay inaad ku koraan goobtan kala duwan ee barashada mashiinka.
