通篇以iris（鸢尾花）数据集为例

（下图为数据集部分内容）

1.如何对其进行数据处理？

从iris数据集中，提取前两个分类的数据，并以[sepal length，sepal width]作为特征，即提取iris数据集中的样品：前100个；变量：第1个，第2个，和倒数第1个。

##载入模块import pandas as pdimport numpy as npfrom sklearn.datasets import load_iris##载入数据iris = load_iris()                                                                    #np的arraydf = pd.DataFrame(iris.data, columns=iris.feature_names)                              #弄成pd的数据框df['label'] = iris.target                                                             #再添加一列作为标签#样品：前100个取过来，变量：第1个，第2个，和倒数第1个。注意第一个下标为0。data = np.array(df.iloc[:100, [0, 1, -1]]) data

2.用sklearn实现感知机

##载入相关模块import numpy as npimport pandas as pdimport matplotlib.pyplot as plt         %matplotlib inlinefrom sklearn.datasets import load_irisfrom sklearn.model_selection import train_test_splitfrom collections import Counter##载入数据iris = load_iris()df = pd.DataFrame(iris.data, columns=iris.feature_names)df['label'] = iris.target      df.columns = ['sepal length', 'sepal width', 'petal length', 'petal width', 'label']##提取特征和样品（取前面100个数，第一列、第二列和最后一列）data = np.array(df.iloc[:100, [0, 1, -1]])      # 最后一个特征作为标签，其他的作为特征X, y = data[:,:-1], data[:,-1]         # 前者表示的是取所有行，但不包括最后一列的数据，结果是个DataFrame。后者则是取所有行最后一列对应的一列数据，结果是Series。         # 取80%作为训练，20%作为测试X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)    ## 载入sklearn中的感知机模块import sklearnfrom sklearn.linear_model import Perceptronclf = Perceptron(fit_intercept=True,                  max_iter=1000,                  shuffle=True)                 clf.fit(X_train, y_train)## 输出感知机参数print(clf.coef_)##画图# 画布大小plt.figure(figsize=(10,10))# 标题plt.rcParams['font.sans-serif']=['SimHei']plt.rcParams['axes.unicode_minus'] = Falseplt.title('鸢尾花线性数据示例')# 散点plt.scatter(data[:50, 0], data[:50, 1], c='b', label='Iris-setosa',)plt.scatter(data[50:100, 0], data[50:100, 1], c='orange', label='Iris-versicolor')# 画感知机的线x_ponits = np.arange(4, 8)y_ = -(clf.coef_[0][0]*x_ponits + clf.intercept_)/clf.coef_[0][1]plt.plot(x_ponits, y_)# 其他部分plt.legend()     # 显示图例plt.grid(False)  # 不显示网格plt.xlabel('sepal length')plt.ylabel('sepal width')plt.legend()

3.用sklearn实现k近邻算法

##载入相关模块import numpy as npimport pandas as pdfrom sklearn.datasets import load_irisfrom sklearn.model_selection import train_test_splitfrom collections import Counter##载入数据iris = load_iris()df = pd.DataFrame(iris.data, columns=iris.feature_names)df['label'] = iris.target      df.columns = ['sepal length', 'sepal width', 'petal length', 'petal width', 'label']##提取特征和样品#取前面100个数，第一列、第二列和最后一列data = np.array(df.iloc[:100, [0, 1, -1]])      #最后一个特征作为标签，其他的作为特征X, y = data[:,:-1], data[:,-1]                  #取80%作为训练，20%作为测试X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)    from sklearn.neighbors import KNeighborsClassifierclf = KNeighborsClassifier()clf.fit(X_train, y_train)## 验证算法精度clf.score(X_test, y_test)##预测某点test_point = [[a, b], [c, d]]                  #a/b/c/d为随机数值（你想预测的数值）clf.predict(test_point)

4.用sklearn实现朴素贝叶斯

##载入相关模块import numpy as npimport pandas as pdfrom sklearn.datasets import load_irisfrom sklearn.model_selection import train_test_splitfrom collections import Counter##载入数据iris = load_iris()df = pd.DataFrame(iris.data, columns=iris.feature_names)df['label'] = iris.target      df.columns = ['sepal length', 'sepal width', 'petal length', 'petal width', 'label']##提取特征和样品#取前面100个数，第一列、第二列和最后一列data = np.array(df.iloc[:100, [0, 1, -1]])      #最后一个特征作为标签，其他的作为特征X, y = data[:,:-1], data[:,-1]                  #取80%作为训练，20%作为测试X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)   ## 载入sklearn中的朴素贝叶斯分类器中的高斯朴素贝叶斯from sklearn.naive_bayes import GaussianNBclf = GaussianNB()clf.fit(X_train, y_train)## 验证算法精度clf.score(X_test, y_test)##预测某些点clf.predict([[ a, b ]])                           #a/b为你想预测的数值

5.决策树

##载入相关模块import numpy as npimport pandas as pdfrom sklearn.datasets import load_irisfrom sklearn.model_selection import train_test_splitfrom collections import Counter##载入数据iris = load_iris()df = pd.DataFrame(iris.data, columns=iris.feature_names)df['label'] = iris.target      df.columns = ['sepal length', 'sepal width', 'petal length', 'petal width', 'label']##提取特征和样品#取前面100个数，第一列、第二列和最后一列data = np.array(df.iloc[:100, [0, 1, -1]])      #最后一个特征作为标签，其他的作为特征X, y = data[:,:-1], data[:,-1]                  #取80%作为训练，20%作为测试X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)    ## 载入sklearn中的决策树模块from sklearn.tree import DecisionTreeClassifierfrom sklearn.tree import export_graphvizimport graphvizclf = DecisionTreeClassifier()clf.fit(X_train, y_train,)## 验证算法精度clf.score(X_test, y_test)##绘制决策树tree_pic = export_graphviz(clf, out_file="mytree.pdf")with open('mytree.pdf') as f:    dot_graph = f.read()graphviz.Source(dot_graph)

6.逻辑回归

##载入相关模块import numpy as npimport pandas as pdimport matplotlib.pyplot as plt%matplotlib inlinefrom sklearn.datasets import load_irisfrom sklearn.model_selection import train_test_splitfrom collections import Counter##载入数据iris = load_iris()df = pd.DataFrame(iris.data, columns=iris.feature_names)df['label'] = iris.target      df.columns = ['sepal length', 'sepal width', 'petal length', 'petal width', 'label']##提取特征和样品#取前面100个数，第一列、第二列和最后一列data = np.array(df.iloc[:100, [0, 1, -1]])      #最后一个特征作为标签，其他的作为特征X, y = data[:,:-1], data[:,-1]                  #取80%作为训练，20%作为测试X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)     ## 载入sklearn中的逻辑回归模块from sklearn.linear_model import LogisticRegression## 模型训练clf = LogisticRegression(max_iter=200)clf.fit(X_train, y_train)## 输出模型参数print(clf.coef_, clf.intercept_)## 验证算法精度clf.score(X_test, y_test)## 绘制x_ponits = np.arange(4, 8)y_ = -(clf.coef_[0][0]*x_ponits + clf.intercept_)/clf.coef_[0][1]plt.plot(x_ponits, y_)plt.plot(X[:50, 0], X[:50, 1], 'bo', color='blue', label='0')plt.plot(X[50:, 0], X[50:, 1], 'bo', color='orange', label='1')plt.xlabel('sepal length')plt.ylabel('sepal width')plt.legend()

7.支持向量机

##载入相关模块import numpy as npimport pandas as pdfrom sklearn.datasets import load_irisfrom sklearn.model_selection import train_test_splitfrom collections import Counter##载入数据iris = load_iris()df = pd.DataFrame(iris.data, columns=iris.feature_names)df['label'] = iris.target      df.columns = ['sepal length', 'sepal width', 'petal length', 'petal width', 'label']##提取特征和样品#取前面100个数，第一列、第二列和最后一列data = np.array(df.iloc[:100, [0, 1, -1]])      #最后一个特征作为标签，其他的作为特征X, y = data[:,:-1], data[:,-1]                  #取80%作为训练，20%作为测试X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)   ## 载入sklearn中的支持向量分类器模块from sklearn.svm import SVC##模型训练clf = SVC()clf.fit(X_train, y_train)## 验证算法精度clf.score(X_test, y_test)

8.AdaBoost算法

##载入相关模块import numpy as npimport pandas as pdfrom sklearn.datasets import load_irisfrom sklearn.model_selection import train_test_splitfrom collections import Counter##载入数据iris = load_iris()df = pd.DataFrame(iris.data, columns=iris.feature_names)df['label'] = iris.target      df.columns = ['sepal length', 'sepal width', 'petal length', 'petal width', 'label']##提取特征和样品#取前面100个数，第一列、第二列和最后一列data = np.array(df.iloc[:100, [0, 1, -1]])      #最后一个特征作为标签，其他的作为特征X, y = data[:,:-1], data[:,-1]                  #取80%作为训练，20%作为测试X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)  ## 载入sklearn中的支持向量分类器模块from sklearn.ensemble import AdaBoostClassifier##模型训练clf = AdaBoostClassifier(n_estimators=100, learning_rate=0.5)clf.fit(X_train, y_train)## 验证算法精度clf.score(X_test, y_test)

9.kmeans聚类

##载入相关模块import mathimport randomimport numpy as npfrom sklearn import datasets,clusterimport matplotlib.pyplot as plt##载入数据，获得标签信息iris = load_iris()gt = iris['target'];gt## 获取属性信息data = iris['data'][:,:2]data## 载入sklearn中的kmeans聚类模块from sklearn.cluster import KMeans##模型训练kmeans = KMeans(n_clusters=3, max_iter=100).fit(data)## 得到类标签gt_labels__ = kmeans.labels_gt_labels__## 得到类中心centers__ = kmeans.cluster_centers_centers__## 绘图及可视化cat1 = data[gt_labels__ == 0]cat2 = data[gt_labels__ == 1]cat3 = data[gt_labels__ == 2]for ix, p in enumerate(centers__):    plt.scatter(p[0], p[1], color='C{}'.format(ix), marker='^', edgecolor='black', s=256)        plt.scatter(cat1[:,0], cat1[:,1], color='green')plt.scatter(cat2[:,0], cat2[:,1], color='red')plt.scatter(cat3[:,0], cat3[:,1], color='blue')plt.title('kmeans using sklearn with k=3')plt.xlim(4, 8)plt.ylim(1, 5)plt.show()     ## 寻找K值from sklearn.cluster import KMeansloss = []for i in range(1, 10):    kmeans = KMeans(n_clusters=i, max_iter=100).fit(data)    loss.append(kmeans.inertia_ / len(data) / 3)plt.title('K with loss')plt.plot(range(1, 10), loss)plt.show()

10.梯度下降实现感知机原理

##载入相关模块import numpy as npimport pandas as pdfrom sklearn.datasets import load_irisfrom sklearn.model_selection import train_test_splitfrom collections import Counterimport matplotlib.pyplot as plt##载入数据iris = load_iris()df = pd.DataFrame(iris.data, columns=iris.feature_names)df['label'] = iris.target      df.columns = ['sepal length', 'sepal width', 'petal length', 'petal width', 'label']##提取特征和样品#取前面100个数，第一列、第二列和最后一列data = np.array(df.iloc[:100, [0, 1, -1]])      #最后一个特征作为标签，其他的作为特征X, y = data[:,:-1], data[:,-1]                          #取80%作为训练，20%作为测试#X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)y = np.array([1 if i == 1 else -1 for i in y])    #把原本取值为0和1的y，调整成-1和1class Model:     #初始化    def __init__(self):        #初始化w，b和学习        self.w = np.ones(len(data[0]) - 1, dtype=np.float32)         #data[0]为第一行的数据len（data[0]=3)这里取两个w权重参数        self.b = 0        self.l_rate = 0.1                                            #学习率为0.1        # self.data = data        #定义线性函数    def lin(self, x, w, b):               y = np.dot(x, w) + b        return y    # 随机梯度下降法    def fit(self, X_train, y_train):        is_wrong = False                                             #判断是否误分类        while not is_wrong:            wrong_count = 0                          for d in range(len(X_train)):                             #取出样例，不断迭代                X = X_train[d]                y = y_train[d]                if y * self.lin(X, self.w, self.b) <= 0:              #根据错误的样本点不断更新w和b值                    self.w = self.w + self.l_rate * (y * X)           #w1=w0+rate*x*y                    self.b = self.b + self.l_rate * y                 #b1=b0+y                    wrong_count += 1                                  #角标（迭代次数—1）            if wrong_count == 0:                                      #直到误分类点为0，则跳出循环                is_wrong = True                                                          return 'Perceptron Model!'    def score(self):        pass##模型训练perceptron = Model()                    #实例化感知机perceptron.fit(X, y)                    #进行训练## 参数估计结果perceptron.w[0],perceptron.w[1],perceptron.b## 绘图x_points = np.linspace(4, 7, 10)                                          #x轴的划分y_ = -(perceptron.w[0] * x_points + perceptron.b) / perceptron.w[1]       plt.plot(x_points, y_)plt.plot(data[:50, 0], data[:50, 1], 'bo', color='blue', label='0')plt.plot(data[50:100, 0], data[50:100, 1], 'bo', color='orange', label='1')plt.xlabel('sepal length')plt.ylabel('sepal width')plt.legend()