clustering_keras.py

import keras
from keras.models import Sequential
from keras.layers import Dense, Input
from keras.utils import to_categorical
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score

# turn the x,y coordinates into a predictors matrix
X = np.array([x, y]).T
# turn the labels into 1-hot encodings
Y = np.array(labels).reshape(-1,1)
Y= to_categorical(Y)

# split the data into training and validation sets
X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=.5, 
                                                    shuffle=True, 
                                                    random_state=999)

# create a network with 2 layers
model = Sequential()
model.add(Dense(50, input_shape=(2,), activation='relu'))
model.add(Dense(3, activation='softmax'))
model.summary()

# train the network
model.compile(optimizer='sgd', loss='categorical_crossentropy', metrics=['accuracy'])
history = model.fit(X_train,Y_train,
                    batch_size=5,
                    epochs=20,
                    shuffle=True,
                    verbose=0,
                    validation_data=(X_test, Y_test))

# perform predictions
Y_test_pred = model_cl.predict_proba(X_test)
Y_train_pred = model_cl.predict_proba(X_train)
Y_test_label = np.argmax(Y_test,axis=1)
Y_train_label = np.argmax(Y_train,axis=1)
Y_test_pred_label = np.argmax(Y_test_pred,axis=1)
Y_train_pred_label = np.argmax(Y_train_pred,axis=1)

# plot predictions
fig, ax = plt.subplots(1,2,figsize=(15,5))
for cloud in clouds:
    i = np.where(Y_test_label==cloud['label'], True, False)
    ax[0].scatter(X_test[i][:,0], X_test[i][:,1], marker=cloud['marker'], color=cloud['color'], label=cloud['label'])
    j = np.where(Y_test_pred_label==cloud['label'], True, False)
    ax[1].scatter(X_test[j][:,0], X_test[j][:,1], marker=cloud['marker'], color=cloud['color'], label=cloud['label'])
ax[0].set_xlabel('x')
ax[0].set_ylabel('y')
ax[0].legend()
ax[0].set_title('test data with original labels')
ax[0].legend();
ax[1].set_xlabel('x')
ax[1].set_ylabel('y')
ax[1].legend()
ax[1].set_title('test data with predicted labels');

# print accuracy
print('Test accuracy: {0:0.2%}'.format(accuracy_score(Y_test_label, Y_test_pred_label)))
print('Training accuracy: {0:0.2%}'.format(accuracy_score(Y_train_label, Y_train_pred_label)))
ax[1].legend();