复习日
仔细回顾一下之前21天的内容，没跟上进度的同学补一下进度。
作业：
自行学习参考如何使用kaggle平台，写下使用注意点，并对下述比赛提交代码
一、数据预处理
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import warnings
from sklearn.model_selection import train_test_split
warnings.filterwarnings("ignore")
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False
data = pd.read_csv('./day22/train.csv')
print(data.info())
print(data.isnull().sum())
for i in data.columns:if data[i].isnull().sum() > 0:if pd.api.types.is_numeric_dtype(data[i]):median_val = data[i].median()data[i].fillna(median_val, inplace=True)print(f"用中位数 {median_val} 填补列：{i}")else:zhongshu = data[i].mode()[0]data[i].fillna(zhongshu, inplace=True)print(f"用众数{zhongshu} 填补列：{i}")data = data.drop(columns=['Name','Ticket', 'Cabin'])
print(data.info())
print(data.isnull().sum())data = pd.get_dummies(data, columns=['Embarked'])
data2 = pd.read_csv('./day22/train.csv')
list_final = []
for i in data.columns:if i not in data2.columns:list_final.append(i)
for i in list_final:data[i] = data[i].astype(int) sex_mapping = {'male': 1,'female': 0,
}
data['Sex'] = data['Sex'].map(sex_mapping)
print(data.info())
print(data.isnull().sum())二、利用随机森林模型进行训练和验证
from sklearn.model_selection import train_test_splitX = data.drop(['Survived'], axis=1)
y = data['Survived']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
print(X_train.shape, X_test.shape, y_train.shape, y_test.shape) # (1382, 6) (346, 6) (1382,) (346,)import numpy as np
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import StratifiedKFold, cross_validate
from sklearn.metrics import make_scorer, accuracy_score, precision_score, recall_score, f1_score, confusion_matrix, classification_report
import time
import warnings
warnings.filterwarnings("ignore")
warnings.filterwarnings("ignore")
print("--- 1. 默认参数随机森林 (训练集 -> 测试集) ---")
import time
start_time = time.time()
rf_model = RandomForestClassifier(random_state=42)
rf_model.fit(X_train, y_train)
rf_pred = rf_model.predict(X_test)
end_time = time.time()print(f"训练与预测耗时: {end_time - start_time:.4f} 秒")
print("\n默认随机森林 在测试集上的分类报告：")
print(classification_report(y_test, rf_pred))
print("默认随机森林 在测试集上的混淆矩阵：")
print(confusion_matrix(y_test, rf_pred))from imblearn.over_sampling import SMOTE
smote = SMOTE(random_state=42)
X_train_smote, y_train_smote = smote.fit_resample(X_train, y_train)print("SMOTE过采样后训练集的形状：", X_train_smote.shape, y_train_smote.shape)print("--- 2. 带权重随机森林 + 交叉验证 (在训练集上进行) ---")counts = np.bincount(y_train)
minority_label = np.argmin(counts)
majority_label = np.argmax(counts)
print(f"训练集中各类别数量: {counts}")
print(f"少数类标签: {minority_label}, 多数类标签: {majority_label}")rf_model_weighted = RandomForestClassifier(random_state=42,class_weight='balanced' # class_weight={minority_label: 10, majority_label: 1} cv_strategy = StratifiedKFold(n_splits=5, shuffle=True, random_state=42)
scoring = {'accuracy': 'accuracy','precision_minority': make_scorer(precision_score, average='macro', zero_division=0),'recall_minority': make_scorer(recall_score, average='macro'),'f1_minority': make_scorer(f1_score, average='macro')
}
print(f"开始进行 {cv_strategy.get_n_splits()} 折交叉验证...")
start_time_cv = time.time()cv_results = cross_validate(estimator=rf_model_weighted,X=X_train_smote,y=y_train_smote,cv=cv_strategy,scoring=scoring,n_jobs=-1, return_train_score=False
)end_time_cv = time.time()
print(f"交叉验证耗时: {end_time_cv - start_time_cv:.4f} 秒")print("\n带权重随机森林 交叉验证平均性能 (基于训练集划分)：")
for metric_name, scores in cv_results.items():if metric_name.startswith('test_'): clean_metric_name = metric_name.split('test_')[1]print(f" 平均 {clean_metric_name}: {np.mean(scores):.4f} (+/- {np.std(scores):.4f})")print("-" * 50)print("--- 3. 训练最终的带权重模型 (整个训练集) 并在测试集上评估 ---")
start_time_final = time.time()
rf_model_weighted_final = RandomForestClassifier(random_state=42,class_weight='balanced'
)
rf_model_weighted_final.fit(X_train_smote, y_train_smote)
rf_pred_weighted = rf_model_weighted_final.predict(X_test)
end_time_final = time.time()print(f"最终带权重模型训练与预测耗时: {end_time_final - start_time_final:.4f} 秒")
print("\n带权重随机森林 在测试集上的分类报告：")
print(classification_report(y_test, rf_pred_weighted))
print("带权重随机森林 在测试集上的混淆矩阵：")
print(confusion_matrix(y_test, rf_pred_weighted))
print("-" * 50)print("性能对比 (测试集上的少数类召回率 Recall):")
recall_default = recall_score(y_test, rf_pred, average='macro')
recall_weighted = recall_score(y_test, rf_pred_weighted, average='macro')
print(f" 默认模型: {recall_default:.4f}")
print(f" 带权重模型: {recall_weighted:.4f}")三、导入测试集并对数据测试
test_data = pd.read_csv('./day22/test.csv')
for i in test_data.columns:if test_data[i].isnull().sum() > 0:if pd.api.types.is_numeric_dtype(test_data[i]):median_val = test_data[i].median()test_data[i].fillna(median_val, inplace=True)print(f"用中位数 {median_val} 填补列：{i}")else:zhongshu = test_data[i].mode()[0]test_data[i].fillna(zhongshu, inplace=True)print(f"用众数{zhongshu} 填补列：{i}")test_data = test_data.drop(columns=['Name','Ticket', 'Cabin'])test_data = pd.get_dummies(test_data, columns=['Embarked'])
data2 = pd.read_csv('./day22/test.csv')
list_final = []
for i in test_data.columns:if i not in data2.columns:list_final.append(i)
for i in list_final:test_data[i] = test_data[i].astype(int) sex_mapping = {'male': 1,'female': 0,
}
test_data['Sex'] = test_data['Sex'].map(sex_mapping)
print(test_data.info())
print(test_data.isnull().sum())rf_pred_weighted = rf_model_weighted_final.predict(test_data) output = pd.DataFrame({'PassengerId': test_data['PassengerId'],'Survived': rf_pred_weighted
})output.to_csv('titanic_predictions.csv', index=False)
)