python归一化

最大最小化

from sklearn import preprocessing
import sklearn.datasets
# Loading Iris Dataset
iris = sklearn.datasets.load_iris()
data = iris.data[:, :4];
# Min-Max Normalization 
min_max_scaler = preprocessing.MinMaxScaler(feature_range=(0.1,0.9))
input_array = min_max_scaler.fit_transform(data);

最值归一化

from sklearn import preprocessing
import numpy as np

X = np.array([[1., -1., 2.],
              [2., 0., 0.],
              [0., 1., -1.]])
min_max_scaler = preprocessing.MinMaxScaler()
X_minMax = min_max_scaler.fit_transform(X)
print(X_minMax)

				[[0.5        0.         1.        ]
				 [1.         0.5        0.33333333]
				 [0.         1.         0.        ]]

标准化

# 归一化
import numpy as np
from sklearn.preprocessing import StandardScaler

X_train = np.array([[0, 3, 7, 1], [4, 5, 9, 8], [8, 2, 7, 7], [6, 3, 4, 5]])
X_test = np.array([[0, 3, 7, 1], [2, 5, 4, 7]])
sc = StandardScaler()
## 估算训练数据中的mu和sigma
sc.fit(X_train)
## 使用训练数据中的mu和sigma对数据进行标准化
X_train_std = sc.transform(X_train)
print(X_train_std)
X_test_std = sc.transform(X_test)
print(X_test_std)

反归一化

参考python归一化和反归一化

from sklearn.preprocessing import MinMaxScaler

mm = MinMaxScaler()

# 归一化
train_label = mm.fit_transform(train_data)

# 反归一化
predict_value = mm.inverse_transform(predict_value)
此外，也可以用 from sklearn.preprocessing import StandardScaler 进行标准化。

更新