SVM不同核函数及其参数对分类性能

王**

导入

%pylab inline

import numpy as np
import matplotlib.pyplot as plt
from sklearn.linear_model import SGDClassifier
from sklearn.datasets.samples_generator import make_blobs
from sklearn.preprocessing import StandardScaler
from matplotlib.colors import ListedColormap
from skimage import io,data

from sklearn.feature_selection import RFE
from sklearn.ensemble import GradientBoostingClassifier, IsolationForest
from sklearn.externals import joblib
from sklearn.model_selection  import train_test_split
import numpy.ma as ma
import os, shutil

Populating the interactive namespace from numpy and matplotlib

读取数据

from skimage.io import imread
img = imread('E:\Hyperspectral_Project\dc.tif')
roi = io.imread('E:\Hyperspectral_Project\protest.tif')

img = np.transpose(img,(1,2,0))#(1280, 307,191)
labels = np.unique(roi[roi > 0])
X =img.reshape(392960,191)
t =img.reshape(392960,191)

Y=roi[:,:,0]
np.unique(Y)

Y=Y.ravel()

print(Y.shape)
print(X.shape)

X = X[Y>0,:]
Y = Y[Y>0]

np.unique(Y)

(392960,)
(392960, 191)

array([  2,   3,  51, 102, 153, 204, 255], dtype=int16)

from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import accuracy_score
from ipywidgets import interact,interact_manual

X_train, X_test, y_train, y_test = train_test_split(
        X,
        Y,
        train_size=0.75,
        random_state= 42,
        stratify=Y)

X_train, X_valid, y_train, y_valid = train_test_split(
        X_train,
        y_train,
        train_size=0.66,
        random_state= 0,
        stratify=y_train)

#训练模型
from sklearn.svm import SVC

clf = SVC(kernel='linear').fit(X_train, y_train)

clf.score(X_train, y_train)

1.0

clf.score(X_valid, y_valid)

0.99921104536489147

y_model = clf.predict(X_test)
accuracy_score(y_test, y_model)

0.99758551307847088

import seaborn as sns

from sklearn.metrics import confusion_matrix

mat = confusion_matrix(y_test, y_model)

sns.heatmap(mat, square=True, annot=True,fmt='d', cbar=False)
plt.xlabel('predicted value')
plt.ylabel('true value');