SelectFromModel(estimator=LinearSVC(penalty='l1'))),
("classification", RandomForestClassifier())])
clf.fit(x, y)
# np.random.RandomState(1) # 1 为种子
# RandomState.rand(d0, d1, ..., dn) #The dimensions of the returned array, should all be positive. If no argument is given a single Python float is returned.
# 多层感知机分类
from sklearn.neural_network import MLPClassifier # 利用了BP
x = [[0, 0], [1, 1]] # 使用 SGD 或 Adam ，训练过程支持在线模式和小批量学习模式
y = [0, 1] # 小数据集下，lbfgs运行更快，表现更好；大数据集，adam更好
clf = MLPClassifier(
solver='lbfgs', alpha=1e-5, hidden_layer_sizes=(5, 2),
random_state=1) # alpha：L2 penalty (regularization term) parameter
clf.fit(x, y)
clf.predict([[2, 2], [-1, -2]])
# 区域中浣熊脸的图片分割
from time import time
from sklearn.cluster import spectral_clustering
import matplotlib.pyplot as plt
import numpy as np
import scipy as sp
from sklearn.feature_extraction import image
from scipy.misc import face
face = face(
gray=True) # face.shape=(768, 1024), type(face)= <class 'numpy.ndarray'>
face = sp.misc.imresize(
face, 0.1) / 255 #face.shape=(76,102),缩小当前尺寸的0.1，除以255，将数组归一化
graph = image.img_to_graph(
face
