dnn sample

diabetes.csv

	import pandas as pd
	import numpy as np
	import keras
	from keras.models import Sequential
	from keras.layers import Dense
	dataset = pd.read_csv("diabetes.csv")
	data = dataset.as_matrix()
	x_train = data[:700, 0:8]
	y_train = data[:700, 8]
	x_test = data[700:, 0:8]
	y_test = data[700:, 8]
	model = Sequential()
	model.add(Dense(12, input_dim = 8, activation = 'relu'))
	model.add(Dense(8, activation = 'relu'))
	model.add(Dense(1, activation = 'sigmoid'))
	model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
	model.fit(x_train, y_train, epochs=1500, batch_size=64)
	scores = model.evaluate(x_test, y_test)
	print("%s: %.2f%%" %(model.metrics_names[1], scores[1]* 100))

view raw dnn_sample01.py hosted with ❤ by GitHub

Epoch 1491/1500
700/700 [==============================] - 0s 24us/step - loss: 0.4035 - acc: 0.7929
Epoch 1492/1500
700/700 [==============================] - 0s 29us/step - loss: 0.4029 - acc: 0.8057
Epoch 1493/1500
700/700 [==============================] - 0s 30us/step - loss: 0.4099 - acc: 0.8014
Epoch 1494/1500
700/700 [==============================] - 0s 25us/step - loss: 0.4171 - acc: 0.7943
Epoch 1495/1500
700/700 [==============================] - 0s 26us/step - loss: 0.4090 - acc: 0.7971
Epoch 1496/1500
700/700 [==============================] - 0s 24us/step - loss: 0.4024 - acc: 0.7971
Epoch 1497/1500
700/700 [==============================] - 0s 29us/step - loss: 0.4024 - acc: 0.8057
Epoch 1498/1500
700/700 [==============================] - 0s 24us/step - loss: 0.4057 - acc: 0.7971
Epoch 1499/1500
700/700 [==============================] - 0s 21us/step - loss: 0.4123 - acc: 0.7971
Epoch 1500/1500
700/700 [==============================] - 0s 23us/step - loss: 0.4016 - acc: 0.7957
68/68 [==============================] - 0s 1ms/step
acc: 79.41%