数学建模--分类模型

逻辑回归

import pandas as pd
import numpy as np

data = pd.read_excel('fruit_data.xlsx', index_col="ID")
data.head()

	mass	width	height	color_score	fruit_name
ID
1	192	8.4	7.3	0.55	apple
2	180	8.0	6.8	0.59	apple
3	176	7.4	7.2	0.60	apple
4	178	7.1	7.8	0.92	apple
5	172	7.4	7.0	0.89	apple

train_data = data.dropna()
train_data['category'] = train_data['fruit_name'].apply(lambda x: 1 if x=='apple' else 0)
train_data.head()

R:\Anaconda\lib\site-packages\ipykernel_launcher.py:2: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

	mass	width	height	color_score	fruit_name	category
ID
1	192	8.4	7.3	0.55	apple	1
2	180	8.0	6.8	0.59	apple	1
3	176	7.4	7.2	0.60	apple	1
4	178	7.1	7.8	0.92	apple	1
5	172	7.4	7.0	0.89	apple	1

test_data = data.loc[data['fruit_name'].isnull() == True]
test_data

	mass	width	height	color_score	fruit_name
ID
39	158	7.1	7.6	0.72	NaN
40	190	7.5	7.9	0.77	NaN
41	189	7.6	7.7	0.77	NaN
42	160	7.9	6.9	0.65	NaN

方法一：sklearn.linear_model.LogisticRegression

from sklearn.linear_model import LogisticRegression

X = train_data.iloc[:,:-2]
y = train_data['category']

LR = LogisticRegression()
LR.fit(X, y)

LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,
                   intercept_scaling=1, l1_ratio=None, max_iter=100,
                   multi_class='auto', n_jobs=None, penalty='l2',
                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,
                   warm_start=False)

print(LR.intercept_)

[4.54213181]

print(LR.coef_)

[[-0.01125145  0.97166531 -1.314372    0.20036824]]

test = test_data.iloc[:,:-1]

# 预测
print(LR.predict(test))
print(LR.predict_proba(test))  # 预测为0和1的概率

[0 0 0 1]
[[0.54530945 0.45469055]
 [0.63120971 0.36879029]
 [0.54143416 0.45856584]
 [0.18555923 0.81444077]]

# 准确率
LR.score(X, y)

0.7105263157894737

方法二：statsmodels（结果与SPSS一致）

import statsmodels.api as sm

X1 = sm.add_constant(X)
lr = sm.Logit(y, X1)
result = lr.fit()
result.summary()

Optimization terminated successfully.
         Current function value: 0.449106
         Iterations 7

Logit Regression Results

Dep. Variable:	category	No. Observations:	38
Model:	Logit	Df Residuals:	33
Method:	MLE	Df Model:	4
Date:	Tue, 12 May 2020	Pseudo R-squ.:	0.3521
Time:	12:30:14	Log-Likelihood:	-17.066
converged:	True	LL-Null:	-26.340
Covariance Type:	nonrobust	LLR p-value:	0.0009644

	coef	std err	z	P>|z|	[0.025	0.975]
const	-7.2016	14.503	-0.497	0.620	-35.627	21.224
mass	-0.0238	0.024	-0.982	0.326	-0.071	0.024
width	4.3068	1.844	2.335	0.020	0.692	7.922
height	-3.7497	1.641	-2.286	0.022	-6.965	-0.534
color_score	9.8913	5.746	1.722	0.085	-1.370	21.152

result.predict(sm.add_constant(test))

ID
39    0.147665
40    0.194533
41    0.446099
42    0.972809
dtype: float64

线性判别分析

from sklearn.discriminant_analysis import LinearDiscriminantAnalysis

lda = LinearDiscriminantAnalysis()
X_r = lda.fit(X, y)
X_r.coef_

array([[-0.03206332,  4.57480239, -2.87678633, 10.50469726]])

X_r.score(X, y)

0.7631578947368421

X_r.predict(test)

array([0, 0, 0, 1], dtype=int64)

X_r.predict(X)

array([1, 1, 0, 0, 1, 1, 0, 1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 0, 0, 0,
       0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1], dtype=int64)

# X_r.predict_proba(X)

多分类

data2 = pd.read_excel('mul_fruit.xlsx')
data2.head()

	ID	mass	width	height	color_score	fruit_name	kind
0	1	192	8.4	7.3	0.55	apple	1.0
1	2	180	8.0	6.8	0.59	apple	1.0
2	3	176	7.4	7.2	0.60	apple	1.0
3	4	178	7.1	7.8	0.92	apple	1.0
4	5	172	7.4	7.0	0.89	apple	1.0

train_data2 = data2.dropna()
test2 = data2.loc[data2['fruit_name'].isnull() == True].iloc[:,1:5]
target_names = train_data2['fruit_name'].unique()

X = train_data2.iloc[:,[1,2,3,4]]
y = train_data2['kind']

lda2 = LinearDiscriminantAnalysis(n_components=2)
X_r2 = lda.fit(X, y)

X_r2.score(X, y)

0.8305084745762712

X_r2.predict(test2)

array([3., 3., 3., 1., 2., 4., 1., 3.])

import matplotlib.pyplot as plt

X_rr = X_r2 = lda.fit(X, y).transform(X)

plt.figure()
colors = ['navy', 'turquoise', 'darkorange', 'blue']
lw = 2

for color, i, target_name in zip(colors, [1, 2, 3, 4], target_names):
    plt.scatter(X_rr[y == i, 0], X_rr[y == i, 1], color=color, alpha=.8, lw=lw,
                label=target_name)
    
plt.legend(loc='best', shadow=False, scatterpoints=1)
plt.title('LDA of FRUITS dataset')

Text(0.5, 1.0, 'LDA of FRUITS dataset')

np.set_printoptions(suppress=True)  # 不使用用科学计数法
# X_r2.predict_proba(X)