def test_load_mnist_one_hot(self):
mnist_mat_train, mnist_label_train \
= dataset.load_mnist("sample_data", "training", one_hot = True)
mnist_mat_test, mnist_label_test \
= dataset.load_mnist("sample_data", "testing", one_hot = True)
self.tlog("mnist train data size : " + str(len(mnist_mat_train)))
self.tlog("mnist test data size : " + str(len(mnist_mat_test)))
def test_load_mnist(self):
mnist_mat_train, mnist_label_train \
= dataset.load_mnist("sample_data", "training", [0,1,2,3,4])
mnist_mat_test, mnist_label_test \
= dataset.load_mnist("sample_data", "testing", [0,1,2,3,4])
self.tlog("mnist train data size : " + str(len(mnist_mat_train)))
self.tlog("mnist test data size : " + str(len(mnist_mat_test)))
def test_process(self):
dg_mat_train, dg_label_train = dataset.load_mnist(
"sample_data", "training")
dg_mat_test, dg_label_test = dataset.load_mnist(
"sample_data", "testing")
knn_digit = KNN(dg_mat_train, dg_label_train, 10, 'euclidean')
error_rate = autotest.eval_predict(knn_digit, dg_mat_test,
dg_label_test, self.logging)
self.tlog("digit predict (with basic knn) error rate :" +
str(error_rate))
def test_process(self):
dg_mat_train, dg_label_train = dataset.load_mnist("sample_data",
"training",
one_hot=True)
dg_mat_test, dg_label_test = dataset.load_mnist("sample_data",
"testing",
one_hot=True)
fnn = FNN(dg_mat_train, dg_label_train, [400, 100])
fnn.fit(lr=0.01, epoch=1000, err_th=0.00001, batch_size=100)
error_rate = autotest.eval_predict(fnn,
dg_mat_test,
dg_label_test,
self.logging,
one_hot=True)
self.tlog("digit predict (with fnn) error rate :" + str(error_rate))
def test_process(self):
dg_mat_train, dg_label_train = dataset.load_mnist("sample_data",
"training",
one_hot=True)
dg_mat_test, dg_label_test = dataset.load_mnist("sample_data",
"testing",
one_hot=True)
logistic_reg = LogisticRegression(dg_mat_train, dg_label_train)
logistic_reg.fit(lr=0.0001, epoch=1000, batch_size=100)
error_rate = autotest.eval_predict(logistic_reg,
dg_mat_test,
dg_label_test,
self.logging,
one_hot=True)
self.tlog("digit predict (with logistic regression) error rate :" +
str(error_rate))
def test_process(self):
dg_mat_train, dg_label_train = dataset.load_mnist("sample_data",
"training",
one_hot=True)
dg_mat_test, dg_label_test = dataset.load_mnist("sample_data",
"testing",
one_hot=True)
svc = SVC(dg_mat_train, dg_label_train)
svc.fit(C=1.5,
toler=0.0001,
epoch=1000,
kernel="RBF",
kernel_params={"gamma": 0.7})
error_rate = autotest.eval_predict(svc,
dg_mat_test,
dg_label_test,
self.logging,
one_hot=True)
self.tlog("digit predict (with svc) error rate :" + str(error_rate))