K 近邻算法（二）

from sklearn import datasets
from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier
#加载鸢尾花数据集
X,y = datasets.load_iris(return_X_y = True)
X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.2)
 
knn_clf = KNeighborsClassifier(n_neighbors=6)
 
knn_clf.fit(X_train,y_train)
y_predict = knn_clf.predict(X_test)
sum(y_predict==y_test)/y_test.shape[0]
# 0.8666666666666667

from sklearn.metrics import accuracy_score
accuracy_score(y_test,y_predict)

from sklearn.model_selection import GridSearchCV
from sklearn.svm import SVC
 
# 定义模型和参数网格
model = SVC()
param_grid = {'C': [0.1, 1, 10], 'kernel': ['linear', 'rbf']}
 
# 创建 GridSearchCV 对象并进行训练
grid_search = GridSearchCV(estimator=model, param_grid=param_grid, cv=5)
grid_search.fit(X_train, y_train)
 
# 获取最佳参数和对应的评分
best_params = grid_search.best_params_
best_score = grid_search.best_score_
 
# 使用最佳参数重新训练模型
best_model = grid_search.best_estimator_
best_model.fit(X_train, y_train)
 
# 在测试集上进行预测
y_pred = best_model.predict(X_test)

from sklearn.model_selection import GridSearchCV
x, y = load_iris(return_X_y=True)
 
x_train, x_test, y_train, y_test = \
    train_test_split(x, y, test_size=0.2, stratify=y, random_state=0)
 
# 创建网格搜索对象
estimator = KNeighborsClassifier()
param_grid = {'n_neighbors': [1, 3, 5, 7]}
estimator = GridSearchCV(estimator, param_grid=param_grid, cv=5, verbose=0)
estimator.fit(x_train, y_train)
 
print('最优参数组合:', estimator.best_params_, '最好得分:', estimator.best_score_)
 
print('测试集准确率:', estimator.score(x_test, y_test))
 
# 最优参数组合: {'n_neighbors': 7} 最好得分: 0.9583333333333334
# 测试集准确率: 1.0

import matplotlib.pyplot as plt
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier
import joblib
from collections import Counter
 
def show_digit(idx):
    
    data = pd.read_csv('手写数字识别.csv')
    if idx < 0 or idx > len(data) - 1:
        return
    x = data.iloc[:, 1:]
    y = data.iloc[:,0]
    print('当前数字的标签为:',y[idx])    # 查看当前数字的数值
 
    
    data_ = x.iloc[idx].values
    # 将数据形状修改为 28*28
    data_ = data_.reshape(28, 28)     # 显示当前数字在数据集的图片
    
    # 显示图像
    plt.imshow(data_)
    plt.show()
 
def train_model():
 
    # 1. 加载手写数字数据集
    data = pd.read_csv('手写数字识别.csv')
    x = data.iloc[:, 1:] / 255
    y = data.iloc[:, 0]
 
    # 2. 打印数据基本信息
    print('数据基本信息:', x.shape)
    print('类别数据比例:', Counter(y))
 
    # 3. 分割数据集
    split_data = train_test_split(x, y, test_size=0.2, stratify=y, random_state=0)
    x_train, x_test, y_train, y_test = split_data
 
    # 4. 模型训练
    estimator = KNeighborsClassifier(n_neighbors=3)
    estimator.fit(x_train, y_train)
 
    # 5. 模型评估
    acc = estimator.score(x_test, y_test)
    print('测试集准确率: %.2f' % acc)
 
    # 6. 模型保存
    joblib.dump(estimator, 'knn.pth')
 
 
def test_model():
    # 读取图片数据
    import matplotlib.pyplot as plt
    import joblib
    img = plt.imread('demo.png') # 对于灰度图像，返回的是一个二维数组，其中每个元素是一个介于0和1之间的浮点数，表示该像素的灰度值
    plt.imshow(img)
    # 加载模型
    knn = joblib.load('knn.pth')
    y_pred = knn.predict(img.reshape(1, -1)) #首先将从图片中读取到的数据 (img) 重塑为一维数组形式，以便给 KNN 分类器进行预测。
    print('您绘制的数字是:', y_pred)
 
 
show_digit(1)
    # 训练模型
train_model()
    # 测试模型
test_model()