机器学习算法源代码

以下是用中文实现的几种常见机器学习算法源代码示例（基于Python和常用库）：

1. 线性回归（使用Scikit-learn）

from sklearn.linear_model import LinearRegression
import numpy as np

# 训练数据
X = np.array([[1], [2], [3], [4]])  # 特征
y = np.array([2, 4, 6, 8])          # 标签

# 创建并训练模型
model = LinearRegression()
model.fit(X, y)

# 预测
prediction = model.predict([[5]])
print("预测结果:", prediction)  # 输出≈10

2. K-近邻（KNN）分类器

from sklearn.neighbors import KNeighborsClassifier
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split

# 加载数据
iris = load_iris()
X, y = iris.data, iris.target

# 划分数据集
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)

# 创建并训练模型
knn = KNeighborsClassifier(n_neighbors=3)
knn.fit(X_train, y_train)

# 评估
accuracy = knn.score(X_test, y_test)
print("准确率:", accuracy)

3. 决策树（纯Python实现）

import numpy as np

class DecisionTree:
    def __init__(self, max_depth=None):
        self.max_depth = max_depth

    def fit(self, X, y, depth=0):
        if depth == self.max_depth or len(np.unique(y)) == 1:
            return np.argmax(np.bincount(y))  # 返回多数类

        best_feat, best_thresh = self.find_best_split(X, y)

        if best_feat is None: 
            return np.argmax(np.bincount(y))

        left_idx = X[:, best_feat] <= best_thresh
        right_idx = X[:, best_feat] > best_thresh

        tree = {}
        tree["特征索引"] = best_feat
        tree["阈值"] = best_thresh
        tree["左子树"] = self.fit(X[left_idx], y[left_idx], depth+1)
        tree["右子树"] = self.fit(X[right_idx], y[right_idx], depth+1)
        return tree

    # 简化版分割查找（实际需计算信息增益）
    def find_best_split(self, X, y):
        best_gain = -1
        best_feat, best_thresh = None, None
        # 此处应实现具体分割逻辑
        return best_feat, best_thresh

# 使用示例
X = np.array([[2.5], [3.0], [3.5], [4.0]])
y = np.array([0, 0, 1, 1])
tree = DecisionTree(max_depth=2).fit(X, y)

4. 神经网络（使用PyTorch）

import torch
import torch.nn as nn

# 定义网络结构
class NeuralNet(nn.Module):
    def __init__(self):
        super().__init__()
        self.fc1 = nn.Linear(4, 10)  # 输入层->隐藏层
        self.fc2 = nn.Linear(10, 3)  # 隐藏层->输出层

    def forward(self, x):
        x = torch.relu(self.fc1(x))
        x = self.fc2(x)
        return x

# 训练流程示例
model = NeuralNet()
criterion = nn.CrossEntropyLoss()
optimizer = torch.optim.Adam(model.parameters())

# 假设有数据
# inputs = torch.tensor(...)
# labels = torch.tensor(...)

# 训练循环
for epoch in range(100):
    optimizer.zero_grad()
    outputs = model(inputs)
    loss = criterion(outputs, labels)
    loss.backward()
    optimizer.step()

5. K-Means聚类（使用Scikit-learn）

from sklearn.cluster import KMeans
import numpy as np

# 生成数据
X = np.array([[1, 2], [1, 4], [4, 2], [4, 4]])

# 创建模型
kmeans = KMeans(n_clusters=2)
kmeans.fit(X)

# 获取结果
print("聚类中心:", kmeans.cluster_centers_)
print("标签分配:", kmeans.labels_)

6. 支持向量机（SVM）

from sklearn.svm import SVC
from sklearn.datasets import make_classification

# 生成数据
X, y = make_classification(n_features=4, random_state=42)

# 创建并训练模型
svm = SVC(kernel='linear')
svm.fit(X, y)

# 预测
print("支持向量数量:", len(svm.support_vectors_))