sklearn iris的k近邻实现预测案例

代码如下：

# iris数据加载
from sklearn import datasets
from sklearn.neighbors import KNeighborsClassifier  # 模型调用
from sklearn.metrics import accuracy_score  # 查看准确度
from sklearn.model_selection import train_test_split  # 数据分离调用
import matplotlib.pyplot as plt

iris = datasets.load_iris()
# x输入数据赋值；y输出数据赋值
x = iris.data
y = iris.target


def basic_test():
    """
    基础测试
    :return: void
    """
    # 展示iris数据
    print(iris.data)
    # 属性名
    print(iris.feature_names)
    # 输出结果
    print(iris.target)
    # 结果含有
    print(iris.target_names)
    # 确认维度
    print(iris.data.shape)
    print(iris.target.shape)
    print(x)
    print(y)


def KNN_example():
    """
    KNN近邻分析
    分类问题测试
    :return: void
    """
    # 创建实例
    knn = KNeighborsClassifier(n_neighbors=1)
    print(knn)
    # 模型训练
    knn.fit(x, y)
    # 预测
    print(knn.predict([[1, 2, 3, 4]]))
    x_test = [[1, 2, 3, 4], [2, 4, 1, 2]]
    print(knn.predict(x_test))
    # 设置一个新的knn进行KNN建模
    knn_5 = KNeighborsClassifier(n_neighbors=5)
    knn_5.fit(x, y)
    print(knn_5.predict(x_test))
    # 确认模型结构
    print(knn_5)


def KNN_assess():
    """
    模型训练与测试
    :return: void
    """
    knn_5 = KNeighborsClassifier(n_neighbors=5)
    knn_5.fit(x, y)
    # 评估数据
    y_pred = knn_5.predict(x)
    print(y_pred)
    print(y_pred.shape)
    # 准确率计算
    print(accuracy_score(y, y_pred))

    # 同样的方法预测K=1时的
    knn_1 = KNeighborsClassifier(n_neighbors=1)
    knn_1.fit(x, y)
    y_pred = knn_1.predict(x)
    print(accuracy_score(y, y_pred))

    # 两种k值预测的准确度进行对比，找到最优解


def data_split():
    """
    数据分离
    :return: void
    """
    print(x.shape)
    print(y.shape)
    x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.4)  # 测试数据占40%
    # 分离后数据集的维度确认
    print(x_train.shape, x_test.shape, y_train.shape, y_test.shape)

    # 分离后数据集的训练与评估
    knn_5_s = KNeighborsClassifier(n_neighbors=5)
    knn_5_s.fit(x_train, y_train)
    y_train_pred = knn_5_s.predict(x_train)
    y_test_pred = knn_5_s.predict(x_test)

    # 分离后模型预测的准确率
    print("5训练：", accuracy_score(y_train, y_train_pred))
    print("5测试新数据：", accuracy_score(y_test, y_test_pred))

    # 同理测试k=1时的
    # 分离后数据集的训练与评估
    knn_1_s = KNeighborsClassifier(n_neighbors=1)
    knn_1_s.fit(x_train, y_train)
    y_train_pred = knn_1_s.predict(x_train)
    y_test_pred = knn_1_s.predict(x_test)

    # 分离后模型预测的准确率
    print("1训练：", accuracy_score(y_train, y_train_pred))
    print("1测试新数据：", accuracy_score(y_test, y_test_pred))

    # 测试K取最优值
    k_range = list(range(1, 26))
    score_train = []
    score_test = []
    for k in k_range:
        knn = KNeighborsClassifier(n_neighbors=k)
        knn.fit(x_train, y_train)
        y_train_pred = knn.predict(x_train)
        y_test_pred = knn.predict(x_test)
        score_train.append(accuracy_score(y_train, y_train_pred))
        score_test.append(accuracy_score(y_test, y_test_pred))
    for k in k_range:
        print(k, "训练集的", score_train[k - 1])
        print(k, "测试新数据", score_test[k - 1])

    # 图形展示
    # 展示k值与训练数据集预测准确率之间的关系
    plt.figure()
    plt.plot(k_range, score_train)
    plt.xlabel("K(KNN model)")
    plt.ylabel("Training Accuracy")

    # 测试集的同理
    plt.figure()
    plt.plot(k_range, score_test)
    plt.xlabel("K(KNN model)")
    plt.ylabel("Testing Accuracy")
    plt.show()

    # 对k=11进行预测
    knn_11 = KNeighborsClassifier(n_neighbors=11)
    knn_11.fit(x_train, y_train)
    print(knn_11.predict([[1, 2, 3, 4]]))

最后绘制出，k与测试数据的准确率图像：