一、欠拟合与过拟合
二、数据分离与混淆矩阵
混淆矩阵(Confusion Matrix)
混淆矩阵，又称为误差矩阵，用于衡量分类算法的准确程度
二、模型优化
四、实战准备
五、酶活性预测，使用数据集T-R-train.csv，测试集T-R-test.csv
#加载数据
import pandas as pd
import numpy as np
data_train = pd.read_csv('T-R-train.csv')
data_train.head()
data_test = pd.read_csv('T-R-test.csv')
data_test.head()
#赋值x,y
x_train = data_train.loc[:,'T']
y_train = data_train.loc[:,'rate']
x_test = data_test.loc[:,'T']
y_test = data_test.loc[:,'rate']
#可视化数据
from matplotlib import pyplot as plt
fig1 = plt.figure()
plt.scatter(x_train,y_train)
#转换成一维数组
x_train = np.array(x_train).reshape(-1,1)
x_test = np.array(x_test).reshape(-1,1)
print(x_train.shape,x_test.shape)
#创建线性回归模型
from sklearn.linear_model import LinearRegression
lr1 = LinearRegression()
lr1.fit(x_train,y_train)
#预测数据
y_train_predict = lr1.predict(x_train)
y_test_predict = lr1.predict(x_test)
#计算R二分数，越接近1，说明模型越好
from sklearn.metrics import r2_score
r2_train = r2_score(y_train,y_train_predict)
r2_test = r2_score(y_test,y_test_predict)
print(r2_train,r2_test)
#可视化模型预测结果
x_range = np.linspace(40,90,300).reshape(-1,1)
y_range_predict = lr1.predict(x_range)
fig2 = plt.figure()
plt.plot(x_range,y_range_predict)
plt.scatter(x_train,y_train)
plt.title('prediction data')
plt.xlabel('temperature')
plt.ylabel('rate')
plt.show()
#多项式模型
from sklearn.preprocessing import PolynomialFeatures
poly2 = PolynomialFeatures(degree=2)
x_2_train = poly2.fit_transform(x_train)
x_2_test = poly2.transform(x_test)
print(x_2_train)
lr2 = LinearRegression()
lr2.fit(x_2_train,y_train)
#预测数据
y_2_train_predict = lr2.predict(x_2_train)
y_2_test_predict = lr2.predict(x_2_test)
#计算R二分数，越接近1，说明模型越好
from sklearn.metrics import r2_score
r2_2_train = r2_score(y_train,y_2_train_predict)
r2_2_test = r2_score(y_test,y_2_test_predict)
print(r2_2_train,r2_2_test)
#可视化模型预测结果
x_2_range = np.linspace(40,90,300).reshape(-1,1)
x_2_range_p = poly2.transform(x_2_range)
y_2_range_predict = lr2.predict(x_2_range_p)
print(x_2_range,y_2_range_predict)
fig4 = plt.figure()
plt.plot(x_2_range,y_2_range_predict)
plt.scatter(x_train,y_train)
plt.scatter(x_test,y_test)
plt.title('prediction data')
plt.xlabel('temperature')
plt.ylabel('rate')
plt.show()
poly5 = PolynomialFeatures(degree=5)
x_5_train = poly5.fit_transform(x_train)
x_5_test = poly5.transform(x_test)
print(x_5_train.shape,x_5_train)
lr5 = LinearRegression()
lr5.fit(x_5_train,y_train)
#预测数据
y_5_train_predict = lr5.predict(x_5_train)
y_5_test_predict = lr5.predict(x_5_test)
#计算R二分数，越接近1，说明模型越好
from sklearn.metrics import r2_score
r2_5_train = r2_score(y_train,y_5_train_predict)
r2_5_test = r2_score(y_test,y_5_test_predict)
print(r2_5_train,r2_5_test)
#可视化模型预测结果
x_5_range = np.linspace(40,90,300).reshape(-1,1)
x_5_range_p = poly5.transform(x_5_range)
y_5_range_predict = lr5.predict(x_5_range_p)
# print(x_5_range,y_5_range_predict)
fig4 = plt.figure()
plt.plot(x_5_range,y_5_range_predict)
plt.scatter(x_train,y_train)
plt.scatter(x_test,y_test)
plt.title('prediction data')
plt.xlabel('temperature')
plt.ylabel('rate')
plt.show()
六、芯片质量好坏分类任务，使用数据集data_class_raw.csv
#加载数据
import pandas as pd
import numpy as np
data = pd.read_csv('data_class_raw.csv')
data.head()
#赋值x,y
x = data.drop(['y'],axis=1)
x1 = data.loc[:,'x1']
x2 = data.loc[:,'x2']
y = data.loc[:,'y']
#可视化数据
from matplotlib import pyplot as plt
fig1 = plt.figure()
bad = plt.scatter(x1[y==0],x2[y==0])
good = plt.scatter(x1[y==1],x2[y==1])
plt.title('raw data')
plt.xlabel('x1')
plt.ylabel('x2')
plt.legend([bad,good],['bad','good'])
plt.show()
#异常检测，剔除异常点
from sklearn.covariance import EllipticEnvelope
ad_model = EllipticEnvelope(contamination=0.02)
ad_model.fit(x[y==0])
y_predict_bad = ad_model.predict(x[y==0])
print(y_predict_bad)
fig2 = plt.figure()
bad = plt.scatter(x1[y==0],x2[y==0])
plt.scatter(x1[y==0][y_predict_bad==-1],x2[y==0][y_predict_bad==-1],marker='x',s=150)
good = plt.scatter(x1[y==1],x2[y==1])
plt.title('raw data')
plt.xlabel('x1')
plt.ylabel('x2')
plt.legend([bad,good],['bad','good'])
plt.show()
#处理剔除异常点后的数据，使用数据集data_class_processed.csv，进行主成分分析（PCA）
import pandas as pd
import numpy as np
data = pd.read_csv('data_class_processed.csv')
#赋值x,y
x = data.drop(['y'],axis=1)
x1 = data.loc[:,'x1']
x2 = data.loc[:,'x2']
y = data.loc[:,'y']
#PCA预处理
from sklearn.preprocessing import StandardScaler#标准化处理
from sklearn.decomposition import PCA#降维
x_norm = StandardScaler().fit_transform(x)
pca = PCA(n_components=2)
x_reduced = pca.fit_transform(x_norm)
#计算各维度标准差比例
var_ratio = pca.explained_variance_ratio_
print(var_ratio)
#可视化标准差比例
fig3 = plt.figure()
plt.bar([1,2],var_ratio)
plt.show()
#数据分离，分离出训练数据集和测试数据集
from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test=train_test_split(x,y,random_state=4,test_size=0.4)
print(x_train.shape,x_test.shape,x.shape)
#训练数据集建立KNN模型，完成分类
from sklearn.neighbors import KNeighborsClassifier
knn_10 = KNeighborsClassifier(n_neighbors=10)
knn_10.fit(x_train,y_train)
y_train_predict = knn_10.predict(x_train)
y_test_predict = knn_10.predict(x_test)
#计算准确率
from sklearn.metrics import accuracy_score
accuracy_trian = accuracy_score(y_train,y_train_predict)
accuracy_test = accuracy_score(y_test,y_test_predict)
print(accuracy_trian,accuracy_test)
#可视化分类边界
xx,yy = np.meshgrid(np.arange(0,10,0.05),np.arange(0,10,0.05))
print(xx,yy)
x_range=np.c_[xx.ravel(),yy.ravel()]
print(x_range.shape,x_range)
y_range_predict= knn_10.predict(x_range)
fig4 = plt.figure()
knn_bad = plt.scatter(x_range[:,0][y_range_predict==0],x_range[:,1][y_range_predict==0])
knn_good = plt.scatter(x_range[:,0][y_range_predict==1],x_range[:,1][y_range_predict==1])
bad = plt.scatter(x1[y==0],x2[y==0])
good = plt.scatter(x1[y==1],x2[y==1])
plt.title('predict result')
plt.xlabel('x1')
plt.ylabel('x2')
plt.legend([bad,good,knn_bad,knn_good],['bad','good','knn_bad','knn_good'])
plt.show()
#计算混淆矩阵，计算准确率
from sklearn.metrics import confusion_matrix
cn = confusion_matrix(y_test,y_test_predict)
print(cn)
TP = cn[1,1]
TN = cn[0,0]
FP = cn[0,1]
FN = cn[1,0]
print(TP,TN,FP,FN)
#准确率：整体样本中，预测正确样本数的比例
accuracy = (TP + TN)/(TP + TN+ FP+FN)
print(accuracy)
#灵敏度（召回率）：正样本中，预测正确的比例
recall = TP/(TP+FN)
print(recall)
#特异度：负样本中，预测正确的比例
specificity=TN/(TN+FP)
print(specificity)
#精确率：预测结果为郑的样本中，预测正确的比例
precision = TP/(TP+FP)
print(precision)
#F1分数，综合精确率和召回率的一个判断标准
f1 = 2*precision*recall/(precision+recall)
print(f1)
#尝试不同的n_neighbors
n = [i for i in range(1,21)]
accuracy_train = []
accuracy_test = []
for i in n:
knn = KNeighborsClassifier(n_neighbors=i)
knn.fit(x_train,y_train)
y_train_predict = knn.predict(x_train)
y_test_predict = knn.predict(x_test)
accuracy_train_i = accuracy_score(y_train,y_train_predict)
accuracy_test_i = accuracy_score(y_test,y_test_predict)
accuracy_train.append(accuracy_train_i)
accuracy_test.append(accuracy_test_i)
print(n)
print(accuracy_train)
print(accuracy_test)
fig5 = plt.figure()
plt.subplot(121)
plt.plot(n,accuracy_train,marker='o')
plt.title('train accuracy vs n_neighbors')
plt.xlabel('n_neighbors')
plt.ylabel('train accuracy')
plt.subplot(122)
plt.plot(n,accuracy_test,marker='o')
plt.title('test accuracy vs n_neighbors')
plt.xlabel('n_neighbors')
plt.ylabel('test accuracy')
plt.show()