def train(dataReader, max_epoch):
n_input = dataReader.num_feature
n_hidden = 2
n_output = 1
eta, batch_size = 0.1, 5
eps = 0.01
hp = HyperParameters2(n_input, n_hidden, n_output, eta, max_epoch, batch_size, eps, NetType.BinaryClassifier, InitialMethod.Xavier)
net = NeuralNet2(hp, "Arc_221_epoch")
#net.LoadResult()
net.train(dataReader, 5, True)
#net.ShowTrainingTrace()
ShowTransformation(net, dataReader, max_epoch)
ShowResult2D(net, dataReader, max_epoch)
def train(n_hidden):
n_input = dataReader.num_feature
n_output = dataReader.num_category
eta, batch_size, max_epoch = 0.1, 10, 10000
eps = 0.01
hp = HyperParameters2(n_input, n_hidden, n_output, eta, max_epoch, batch_size, eps, NetType.MultipleClassifier, InitialMethod.Xavier)
net = NeuralNet2(hp, "Bank_2N3")
net.train(dataReader, 100, True)
net.ShowTrainingTrace()
loss = net.GetLatestAverageLoss()
fig = plt.figure(figsize=(6,6))
ShowDataByOneHot2D(dataReader.XTrain[:,0], dataReader.XTrain[:,1], dataReader.YTrain, hp.toString())
ShowClassificationResult25D(net, 50, str.format("{0}, loss={1:.3f}", hp.toString(), loss))
plt.show()
# draw train data
X, Y = dataReader.XTrain, dataReader.YTrain
plt.plot(X[:, 0], Y[:, 0], '.', c='b')
# create and draw visualized validation data
TX = np.linspace(0, 1, 100).reshape(100, 1)
TY = net.inference(TX)
plt.plot(TX, TY, 'x', c='r')
plt.title(title)
plt.show()
#end def
if __name__ == '__main__':
dataReader = DataReader(train_data_name, test_data_name)
dataReader.ReadData()
dataReader.GenerateValidationSet()
n_input, n_hidden, n_output = 1, 4, 1
eta, batch_size, max_epoch = 0.7, 15, 10000
eps = 0.001
hp = HyperParameters2(n_input, n_hidden, n_output, eta, max_epoch,
batch_size, eps, NetType.Fitting,
InitialMethod.Xavier)
net = NeuralNet2(hp, "complex_121")
net.train(dataReader, 50, True)
net.ShowTrainingTrace()
ShowResult(net, dataReader, hp.toString())
def train(hp, folder):
net = NeuralNet2(hp, folder)
net.train(dataReader, 50, True)
trace = net.GetTrainingTrace()
return trace
p2, = plt.plot(TX, net.A1[:, 1], '.', c='g')
p3, = plt.plot(TX, net.Z2[:, 0], '.', c='blue')
plt.legend([p1, p2, p3], ["a1", "a2", "z"])
plt.show()
if __name__ == '__main__':
dataReader = DataReader(x_data_name, y_data_name)
dataReader.ReadData()
dataReader.GenerateValidationSet()
n_input, n_hidden, n_output = 1, 2, 1
eta, batch_size, max_epoch = 0.05, 10, 5000
eps = 0.001
hp = HyperParameters2(n_input, n_hidden, n_output, eta, max_epoch,
batch_size, eps, NetType.Fitting,
InitialMethod.Xavier)
net = NeuralNet2(hp, "sin_121")
net.LoadResult()
print(net.wb1.W)
print(net.wb1.B)
print(net.wb2.W)
print(net.wb2.B)
#net.train(dataReader, 50, True)
#net.ShowTrainingTrace()
#ShowResult(net, dataReader, hp.toString())
ShowResult2D(net, hp.toString())
# Copyright (c) Microsoft. All rights reserved.
# Licensed under the MIT license. See LICENSE file in the project root for full license information.
import numpy as np
from HelperClass2.DataReader import *
from HelperClass2.HyperParameters2 import *
from HelperClass2.NeuralNet2 import *
train_data_name = "../../Data/ch10.train.npz"
test_data_name = "../../Data/ch10.test.npz"
if __name__ == '__main__':
dataReader = DataReader(train_data_name, test_data_name)
dataReader.ReadData()
dataReader.NormalizeX()
dataReader.Shuffle()
dataReader.GenerateValidationSet()
n_input = dataReader.num_feature
n_hidden = 2
n_output = 1
eta, batch_size, max_epoch = 0.1, 5, 10000
eps = 0.08
hp = HyperParameters2(n_input, n_hidden, n_output, eta, max_epoch, batch_size, eps, NetType.BinaryClassifier, InitialMethod.Xavier)
net = NeuralNet2(hp, "Arc_221")
net.train(dataReader, 5, True)
net.ShowTrainingTrace()
for i in range(count):
for j in range(Y.shape[1]):
if Y[i,j] == 1:
ax.scatter(net.A1[i,0],net.A1[i,1],net.A1[i,2], color=colors[j], marker=shapes[j])
plt.show()
if __name__ == '__main__':
dataReader = DataReader(train_data_name, test_data_name)
dataReader.ReadData()
dataReader.NormalizeY(YNormalizationMethod.MultipleClassifier, base=1)
dataReader.NormalizeX()
dataReader.Shuffle()
dataReader.GenerateValidationSet()
n_input = dataReader.num_feature
n_hidden = 3
n_output = dataReader.num_category
eta, batch_size, max_epoch = 0.1, 10, 5000
eps = 0.1
hp = HyperParameters2(n_input, n_hidden, n_output, eta, max_epoch, batch_size, eps, NetType.MultipleClassifier, InitialMethod.Xavier)
net = NeuralNet2(hp, "Bank_233_2")
#net.LoadResult()
net.train(dataReader, 100, True)
net.ShowTrainingTrace()
Show3D(net, dataReader)
A2 = net.inference(X)
print("A2=", A2)
diff = np.abs(A2 - Y)
result = np.where(diff < 1e-2, True, False)
if result.sum() == dataReader.num_test:
return True
else:
return False
if __name__ == '__main__':
dataReader = XOR_DataReader()
dataReader.ReadData()
dataReader.GenerateValidationSet()
n_input = dataReader.num_feature
n_hidden = 2
n_output = 1
eta, batch_size, max_epoch = 0.1, 1, 10000
eps = 0.005
hp = HyperParameters2(n_input, n_hidden, n_output, eta, max_epoch,
batch_size, eps, NetType.BinaryClassifier,
InitialMethod.Xavier)
net = NeuralNet2(hp, "Xor_221")
net.train(dataReader, 100, True)
net.ShowTrainingTrace()
print(Test(dataReader, net))
