第九章：AI大模型的产业应用与前景9.1 产业应用案例9.1.1 金融

    import numpy as np
    import pandas as pd
    from sklearn.linear_model import LogisticRegression
    from sklearn.model_selection import train_test_split
    from sklearn.metrics import accuracy_score
    
    # 加载数据
    data = pd.read_csv('loan.csv')
    
    # 数据预处理
    data = data.fillna(0)
    data = pd.get_dummies(data)
    
    # 特征选择
    features = data.drop('loan_status', axis=1)
    labels = data['loan_status']
    
    # 训练模型
    model = LogisticRegression()
    model.fit(features, labels)
    
    # 评估模型
    X_test = features.iloc[:1000]
    y_test = labels.iloc[:1000]
    predictions = model.predict(X_test)
    accuracy = accuracy_score(y_test, predictions)
    
    print('Accuracy:', accuracy)
    
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
    
    代码解读

    import numpy as np
    
    # 数据
    X = np.array([[1, 2], [2, 3], [3, 4], [4, 5]])
    y = np.array([0, 1, 0, 1])
    
    # 参数
    theta = np.zeros(X.shape[1])
    learning_rate = 0.01
    iterations = 1000
    
    # 梯度下降
    for i in range(iterations):
    predictions = X.dot(theta)
    loss = np.sum(y * np.log(predictions) + (1 - y) * np.log(1 - predictions)) / X.shape[0]
    gradient = np.dot(X.T, (y - predictions)) / X.shape[0]
    theta = theta - learning_rate * gradient
    
    print('theta:', theta)
    
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
    
    代码解读

    import numpy as np
    import pandas as pd
    from statsmodels.tsa.ar import AutoReg
    
    # 加载数据
    data = pd.read_csv('stock.csv')
    
    # 数据预处理
    data['date'] = pd.to_datetime(data['date'])
    data.set_index('date', inplace=True)
    
    # 特征选择
    features = data['close']
    labels = data['volume']
    
    # 训练模型
    model = AutoReg(features, lags='2')
    model_fit = model.fit()
    
    # 评估模型
    predictions = model_fit.predict(start='2010-01-01', end='2020-12-31')
    
    print('Predictions:', predictions)
    
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
    
    代码解读

    import numpy as np
    
    # 数据
    X = np.array([1, 2, 3, 4, 5])
    y = np.array([2, 3, 3, 4, 5])
    
    # 参数
    phi = np.array([0.5, 0.3, 0.2, 0.1, 0.1])
    epsilon = np.array([0.1, 0.2, 0.1, 0.3, 0.4])
    
    # 自回归
    predictions = np.zeros(X.shape[0])
    for i in range(X.shape[0]):
    predictions[i] = phi[0] * X[i] + phi[1] * X[i - 1] + phi[2] * X[i - 2] + phi[3] * X[i - 3] + phi[4] * X[i - 4] + epsilon[i]
    
    print('Predictions:', predictions)
    
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
    
    代码解读

    import numpy as np
    import pandas as pd
    from sklearn.ensemble import IsolationForest
    
    # 加载数据
    data = pd.read_csv('fraud.csv')
    
    # 数据预处理
    data = data.fillna(0)
    
    # 特征选择
    features = data.drop('fraud_label', axis=1)
    labels = data['fraud_label']
    
    # 训练模型
    model = IsolationForest(n_estimators=100, max_samples='auto', contamination=0.01, random_state=42)
    model.fit(features)
    
    # 评估模型
    X_test = features.iloc[:1000]
    y_test = labels.iloc[:1000]
    predictions = model.predict(X_test)
    
    print('Predictions:', predictions)
    
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
    
    代码解读

    import numpy as np
    
    # 数据
    X = np.array([[1, 2], [2, 3], [3, 4], [4, 5], [5, 6]])
    labels = np.array([0, 0, 0, 1, 1])
    
    # 参数
    n_estimators = 100
    max_samples = 'auto'
    contamination = 0.01
    random_state = 42
    
    # 异常检测
    model = IsolationForest(n_estimators=n_estimators, max_samples=max_samples, contamination=contamination, random_state=random_state)
    model.fit(X)
    
    # 预测
    predictions = model.predict(X)
    
    print('Predictions:', predictions)
    
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
    
    代码解读

    import numpy as np
    import pandas as pd
    from scipy.sparse import csr_matrix
    from scipy.sparse.linalg import svds
    
    # 加载数据
    data = pd.read_csv('user_behavior.csv')
    
    # 数据预处理
    data = data.fillna(0)
    
    # 特征选择
    users = data['user_id'].unique()
    items = data['item_id'].unique()
    
    # 构建用户-项矩阵
    user_item_matrix = csr_matrix((data['rating'].values, (data['user_id'].values, data['item_id'].values)), shape=(len(users), len(items)))
    
    # 协同过滤
    U, s, Vt = svds(user_item_matrix, k=10)
    predictions = U.dot(s).dot(Vt)
    
    print('Predictions:', predictions)
    
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
      
    
    代码解读

    import numpy as np
    
    # 数据
    X = np.array([[1, 2, 3], [2, 3, 4], [3, 4, 5], [4, 5, 6]])
    ratings = np.array([[3, 2, 1], [2, 3, 4], [1, 4, 5], [4, 5, 6]])
    
    # 参数
    k = 10
    
    # 协同过滤
    U, s, Vt = svds(ratings, k=k)
    predictions = U.dot(s).dot(Vt)
    
    print('Predictions:', predictions)
    
      
      
      
      
      
      
      
      
      
      
      
      
      
    
    代码解读