Ukwakha imodeli yenethiwekhi ye-neural kuwumkhakha othakazelisayo ekufundeni komshini, ikakhulukazi ePython. Inikeza ububanzi obuningi bokuhlaziya, ukuqagela, kanye nezinqubo zokuthatha izinqumo ezizenzakalelayo. Ngaphambi kokuthi singene ku-nitty-gritty yokwakha inethiwekhi ye-neural yesakhiwo, kubalulekile ukuqonda ukuthi iyini inethiwekhi ye-neural. Empeleni kuwuhlelo lwama-algorithms olusondeza ukwakheka kobuchopho bomuntu, ngaleyo ndlela kwakheka inethiwekhi ye-neural yokwenziwa okuthi, ngenqubo yokuhlaziya ihumushe idatha yezinzwa, icoshe ama-nuances 'angabonakali' ngedatha eluhlaza, njengoba kwenza ubuchopho bethu.
Inethiwekhi ye-neural ibalulekile ezinqubweni zokumbiwa kwedatha, lapho ihlonza amaphethini namathrendi abeyinkimbinkimbi kakhulu kubantu noma amanye amasu ekhompyutha. Manje, ake singene enhliziyweni yendabaโ sisebenzisa iPython ukwakha nokuhlela inethiwekhi ye-neural.
Ukwakha amanethiwekhi we-neural ePython
# 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()
Masiqonde le khodi:
- Emigqeni emine yokuqala, singenisa imitapo yolwazi edingekayo njenge-numpy, matplotlib njll.
- Okulandelayo, sisebenzisa umsebenzi we-'make_blobs' kusuka ku-sklearn, sakha idathasethi.
- Bese isethi yedatha ihlukaniswa okokufaka (X) kanye nokuphumayo (y).
- Umugqa wokugcina uhlela u-X kanye no-y futhi usinika umbono wedatha kusetshenziswa umsebenzi we-scatter kusuka kulabhulali ye-matplotlib.
Ukuqonda imitapo yolwazi yenethiwekhi ye-neural
Ukuqonda imitapo yolwazi yePython kulo mongo kubalulekile. Umtapo wezincwadi we-numpy uvumela ukusebenza kwezibalo, i-matplotlib isetshenziselwa ukuhlela igrafu ye-2D kusukela kudatha eku-Python kanye nokufunda komshini we-sklearn spearheads ku-Python.
Ikhodi yesinyathelo ngesinyathelo
Isinyathelo ngesinyathelo inqubo yekhodi isivumela ukuthi sithole ukuqonda okujulile:
# 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)
Kulesi siqephu sekhodi,
- Sakha imodeli sisebenzisa umsebenzi weSequential() kusuka kumojula ye-keras.models.
- Okulandelayo, isendlalelo sokufakwayo singezwa ngama-neurons okufakwayo angu-2. Lapha, i-'Dense' iwuhlobo lwesendlalelo olusebenza ezimweni eziningi. Kusendlalelo esiminyene, wonke ama-node akusendlalelo sangaphambilini axhuma kuma-node kusendlalelo samanje.
- 'Ukuhlanganisa' kulungiselela imodeli yokuqeqeshwa.
- Ingxenye yokugcina, 'ukufaka imodeli' yilapho inethiwekhi ye-neural iqeqeshwa khona. 'Ama-Epoch' abonisa inani lokuphasa kwedathasethi yokuqeqeshwa. Imodeli ifunda futhi ibuyekeze amapharamitha emodeli phakathi nenkathi ngayinye. Usayizi wenqwaba uyisethi engaphansi yedathasethi.
Ngala makhodi, sakha isisekelo sokwakha inethiwekhi ye-neural yesakhiwo sisebenzisa iPython. Ngemitapo yolwazi ebanzi ye-Python namandla anamandla, amanethiwekhi we-neural angasetshenziswa futhi abonwe ngempumelelo. Kumayelana nokuqonda izimpande, futhi kuhle ukuthi ukhule kulo mkhakha oguquguqukayo wokufunda ngomshini.