Kaggle-Predict Calorie Expenditure-(回归+xgb+cat+lgb+模型融合+预测结果)

Predict Calorie Expenditure

题意：

给出每个人的基本信息，预测运动后的卡路里消耗值。

数据处理：

1.构造出人体机能、运动相关的特征值。
2.所有特征值进行从新组合，注意唯独爆炸
3.对连续信息分箱变成离散

建立模型：

1.xgb模型，lgb模型，cat模型
2.使用stack堆叠融合，使用3折交叉验证
3.对xgb、lgb、cat进行K折交叉验证，最终和stack进行结果融合。

代码：
import os
import sys
import warnings
import numpy as np
import pandas as pd
import seaborn
from catboost import CatBoostRegressor
from lightgbm import LGBMRegressor
from matplotlib import pyplot as plt
import lightgbm
from mlxtend.regressor import StackingCVRegressor
from sklearn import clone
from sklearn.ensemble import VotingRegressor, StackingClassifier, StackingRegressor
from sklearn.linear_model import Lasso, LogisticRegression, RidgeCV
from sklearn.metrics import mean_squared_error, mean_absolute_error, r2_score, make_scorer, mean_squared_log_error
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.preprocessing import StandardScaler
from xgboost import XGBRegressor
from sklearn.preprocessing import RobustScaler
from sklearn.model_selection import KFold
from sklearn.linear_model import Ridgedef init():os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'  # 仅输出错误日志warnings.simplefilter('ignore')  # 忽略警告日志pd.set_option('display.width', 1000)pd.set_option('display.max_colwidth', 1000)pd.set_option("display.max_rows", 1000)pd.set_option("display.max_columns", 1000)def show_dataframe(df):print("查看特征值和特征值类型\n" + str(df.dtypes) + "\n" + "-" * 100)print("查看前10行信息\n" + str(df.head()) + "\n" + "-" * 100)print("查看每个特征值的各种数据统计信息\n" + str(df.describe()) + "\n" + "-" * 100)print("输出重复行的个数\n" + str(df.duplicated().sum()) + "\n" + "-" * 100)print("查看每列的缺失值个数\n" + str(df.isnull().sum()) + "\n" + "-" * 100)print("查看缺失值的具体信息\n" + str(df.info()) + "\n" + "-" * 100)#print("输出X所有值出现的是什么，还有对应出现的次数\n" + str(df['X'].value_counts()) + "\n" + "-" * 100)def show_relation(data, colx, coly):  # 输出某一特征值与目标值的关系if data[colx].dtype == 'object' or data[colx].dtype == 'category' or len(data[colx].unique()) < 20:seaborn.boxplot(x=colx, y=coly, data=data)else:plt.scatter(data[colx], data[coly])plt.xlabel(colx)plt.ylabel(coly)plt.show()# 自定义RMSLE评分函数（GridSearchCV需要最大化评分，因此返回负RMSLE）
def rmsle_scorer(y_true, y_pred):y_pred = np.clip(y_pred, 1e-15, None)  # 防止对0取对数y_true = np.clip(y_true, 1e-15, None)log_error = np.log(y_pred + 1) - np.log(y_true + 1)rmsle = np.sqrt(np.mean(log_error ** 2))return -rmsle  # 返回负值，因为GridSearchCV默认最大化评分if __name__ == '__main__':init()df_train = pd.read_csv('/kaggle/input/playground-series-s5e5/train.csv')df_test = pd.read_csv('/kaggle/input/playground-series-s5e5/test.csv')#for col in df_train.columns:#   show_relation(df_train, col, 'Calories')#特征工程df_all = pd.concat([df_train.drop(['id', 'Calories'], axis=1), df_test.drop(['id'], axis=1)], axis=0)df_all['Sex'] = df_all['Sex'].map({'male': 0, 'female': 1})df_all = df_all.reset_index(drop=True)#构造BMIdf_all['BMI'] = df_all['Weight'] / (df_all['Height'] / 100) ** 2#Harris-Benedict公式df_all['BMR'] = 0df_all.loc[df_all['Sex'] == 0, 'BMR'] = 88.362 + (13.397 * df_all['Weight']) + (4.799 * df_all['Height']) - (5.677 * df_all['Age'])df_all.loc[df_all['Sex'] == 1, 'BMR'] = 447.593 + (9.247 * df_all['Weight']) + (3.098 * df_all['Height']) - (4.330 * df_all['Age'])# 数值特征标准化#numeric_features = ['Age', 'Height', 'Weight', 'Duration', 'Heart_Rate', 'Body_Temp']#scaler = StandardScaler()#df_all[numeric_features] = scaler.fit_transform(df_all[numeric_features])#运动强度特征df_all['Max_HR'] = 220 - df_all['Age']  # 最大心率df_all['HR_Reserve_Ratio'] = df_all['Heart_Rate'] / df_all['Max_HR']#交互特征df_all['Weight_Duration'] = df_all['Weight'] * df_all['Duration']df_all['Sex_Weight'] = df_all['Sex'] * df_all['Weight']# 构造运动功率特征df_all['workload'] = df_all['Weight'] * df_all['Duration'] * df_all['Heart_Rate'] / 1000# 构造生理特征交互项df_all['age_heart_ratio'] = df_all['Age'] / df_all['Heart_Rate']# 时间维度特征（如有时间戳）df_all['hour_of_day'] = df_all['Duration']/60/24# 组合特征numeric_cols = df_all.columnsfor i in range(len(numeric_cols)):feature_1 = numeric_cols[i]for j in range(i + 1, len(numeric_cols)):feature_2 = numeric_cols[j]df_all[f'{feature_1}_x_{feature_2}'] = df_all[feature_1] * df_all[feature_2]#数值归一化#scaler = RobustScaler()#df_all = scaler.fit_transform(df_all)now_col = ['Age', 'Height', 'Weight', 'Duration', 'Heart_Rate', 'Body_Temp', 'BMI']for i in now_col:df_all[i + "_box"] = pd.cut(df_all[i], bins=10, labels=False, right=False)X_train = df_all[:df_train.shape[0]]Y_train = np.log1p(df_train['Calories'])x_test = df_all[df_train.shape[0]:]#xgbmodel_xgb =estimator=XGBRegressor(random_state=42,n_estimators=8000,objective='reg:squarederror',eval_metric='rmse',device='cuda',learning_rate=0.05,max_depth=8,colsample_bytree=0.75,subsample=0.9,#reg_lambda = 1,#reg_alpha = 0.5,early_stopping_rounds=500,)#lgbmodel_lgb = lightgbm.LGBMRegressor(n_estimators=3000,  # 增加迭代次数配合早停learning_rate=0.03,  # 减小学习率num_leaves=15,  # 限制模型复杂度min_child_samples=25,  # 增加最小叶子样本数reg_alpha=0.1,  # L1正则化reg_lambda=0.1,  # L2正则化objective='regression_l1',  # 改用MAE损失early_stopping_rounds=500,)#catmodel_cat = CatBoostRegressor(iterations=3500,learning_rate=0.02,depth=12,loss_function='RMSE',l2_leaf_reg=3,random_seed=42,eval_metric='RMSE',early_stopping_rounds=200,verbose=1000,task_type='GPU',)#融合#创建基模型列表（需禁用早停以生成完整预测）base_models = [('xgb', XGBRegressor(early_stopping_rounds=None,  # 禁用早停**{k: v for k, v in model_xgb.get_params().items() if k != 'early_stopping_rounds'})),('lgb', LGBMRegressor(early_stopping_rounds=None,  # 禁用早停**{k: v for k, v in model_lgb.get_params().items() if k != 'early_stopping_rounds'})),('cat', CatBoostRegressor(early_stopping_rounds=None,  # 禁用早停**{k: v for k, v in model_cat.get_params().items() if k != 'early_stopping_rounds'}))]meta_model = RidgeCV()model_stack = StackingRegressor(estimators=base_models,final_estimator=meta_model,cv=3,  # 使用3折交叉验证生成元特征passthrough=False,  # 不使用原始特征verbose=1)FOLDS = 20KF = KFold(n_splits=FOLDS, shuffle=True, random_state=42)cat_features = ['Sex']oof_cat = np.zeros(len(df_train))pred_cat = np.zeros(len(df_test))oof_xgb = np.zeros(len(df_train))pred_xgb = np.zeros(len(df_test))oof_lgb = np.zeros(len(df_train))pred_lgb = np.zeros(len(df_test))for i, (train_idx, valid_idx) in enumerate(KF.split(X_train, Y_train)):print('#' * 15, i + 1, '#' * 15)## SPLIT DSx_train, y_train = X_train.iloc[train_idx], Y_train.iloc[train_idx]x_valid, y_valid = X_train.iloc[valid_idx], Y_train.iloc[valid_idx]## CATBOOST fitmodel_cat.fit(x_train, y_train, eval_set=[(x_valid, y_valid)], cat_features=cat_features,use_best_model=True, verbose=0)## XGB FIRmodel_xgb.fit(x_train, y_train, eval_set=[(x_valid, y_valid)], verbose=0)## LGB MODELmodel_lgb.fit(x_train, y_train, eval_set=[(x_valid, y_valid)])## PREDICTION CATBOOSToof_cat[valid_idx] = model_cat.predict(x_valid)pred_cat += model_cat.predict(x_test)## PREDICTION XGBoof_xgb[valid_idx] = model_xgb.predict(x_valid)pred_xgb += model_xgb.predict(x_test)## PREDICTION LGBoof_lgb[valid_idx] = model_lgb.predict(x_valid)pred_lgb += model_lgb.predict(x_test)cat_rmse = mean_squared_error(y_valid, oof_cat[valid_idx]) ** 0.5xgb_rmse = mean_squared_error(y_valid, oof_xgb[valid_idx]) ** 0.5lgb_rmse = mean_squared_error(y_valid, oof_lgb[valid_idx]) ** 0.5print(f'FOLD {i + 1} CATBOOST_RMSE = {cat_rmse:.4f} <=> XGB_RMSE = {xgb_rmse:.4f} <=> LGB_RMSE = {lgb_rmse:.4f}')#预测pred_cat /= FOLDSpred_xgb /= FOLDSpred_lgb /= FOLDSpred_stack = model_stack.predict(df_test)pred_all = np.expm1(pred_xgb) * 0.1 + np.expm1(pred_stack) * 0.80 + np.expm1(pred_cat) * 0.1submission = pd.DataFrame({'id': df_test['id'],'Calories': pred_all})submission['Calories'] = np.clip(submission['Calories'], a_min=1, a_max=20*df_test['Duration'])submission.to_csv('/kaggle/working/submission.csv', index=False)
代码

使用k折交叉验证，对预测结果再进行训练预测。

import os
import sys
import warnings
import numpy as np
import pandas as pd
import seaborn
from catboost import CatBoostRegressor
from lightgbm import LGBMRegressor
from matplotlib import pyplot as plt
import lightgbm
from mlxtend.regressor import StackingCVRegressor
from sklearn import clone
from sklearn.ensemble import VotingRegressor, StackingClassifier, StackingRegressor
from sklearn.linear_model import Lasso, LogisticRegression, RidgeCV
from sklearn.metrics import mean_squared_error, mean_absolute_error, r2_score, make_scorer, mean_squared_log_error
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.preprocessing import StandardScaler
from xgboost import XGBRegressor
from sklearn.preprocessing import RobustScaler
from sklearn.model_selection import KFold
from sklearn.linear_model import Ridgedef init():os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'  # 仅输出错误日志warnings.simplefilter('ignore')  # 忽略警告日志pd.set_option('display.width', 1000)pd.set_option('display.max_colwidth', 1000)pd.set_option("display.max_rows", 1000)pd.set_option("display.max_columns", 1000)def show_dataframe(df):print("查看特征值和特征值类型\n" + str(df.dtypes) + "\n" + "-" * 100)print("查看前10行信息\n" + str(df.head()) + "\n" + "-" * 100)print("查看每个特征值的各种数据统计信息\n" + str(df.describe()) + "\n" + "-" * 100)print("输出重复行的个数\n" + str(df.duplicated().sum()) + "\n" + "-" * 100)print("查看每列的缺失值个数\n" + str(df.isnull().sum()) + "\n" + "-" * 100)print("查看缺失值的具体信息\n" + str(df.info()) + "\n" + "-" * 100)#print("输出X所有值出现的是什么，还有对应出现的次数\n" + str(df['X'].value_counts()) + "\n" + "-" * 100)def show_relation(data, colx, coly):  # 输出某一特征值与目标值的关系if data[colx].dtype == 'object' or data[colx].dtype == 'category' or len(data[colx].unique()) < 20:seaborn.boxplot(x=colx, y=coly, data=data)else:plt.scatter(data[colx], data[coly])plt.xlabel(colx)plt.ylabel(coly)plt.show()def show_score(model_name, pred):mse = mean_squared_error(y_val, pred)mae = mean_absolute_error(y_val, pred)score = r2_score(y_val, pred)print(model_name)print(f"{'MSE':<10}{mse:<15.4f}")print(f"{'MAE':<10}{mae:<15.4f}")print(f"{'R²':<10}{score:<15.4f}")print("-" * 100)# Function to calculate RMSLE
def rmsle(y_true, y_pred):return np.sqrt(np.mean(np.power(np.log1p(y_true) - np.log1p(y_pred), 2)))if __name__ == '__main__':init()df_train = pd.read_csv('train.csv')df_test = pd.read_csv('test.csv')#for col in df_train.columns:#   show_relation(df_train, col, 'Calories')#特征工程df_all = pd.concat([df_train.drop(['id', 'Calories'], axis=1), df_test.drop(['id'], axis=1)], axis=0)df_all['Sex_encoded'] = df_all['Sex'].map({'male': 0, 'female': 1})df_all.drop(['Sex'], axis=1, inplace=True)df_all = df_all.reset_index(drop=True)#构造BMIdf_all['BMI'] = df_all['Weight'] / (df_all['Height'] / 100) ** 2#Harris-Benedict公式df_all['BMR'] = 0df_all.loc[df_all['Sex_encoded'] == 0, 'BMR'] = 88.362 + (13.397 * df_all['Weight']) + (4.799 * df_all['Height']) - (5.677 * df_all['Age'])df_all.loc[df_all['Sex_encoded'] == 1, 'BMR'] = 447.593 + (9.247 * df_all['Weight']) + (3.098 * df_all['Height']) - (4.330 * df_all['Age'])# 数值特征标准化#numeric_features = ['Age', 'Height', 'Weight', 'Duration', 'Heart_Rate', 'Body_Temp']#scaler = StandardScaler()#df_all[numeric_features] = scaler.fit_transform(df_all[numeric_features])#运动强度特征df_all['Max_HR'] = 220 - df_all['Age']  # 最大心率df_all['HR_Reserve_Ratio'] = df_all['Heart_Rate'] / df_all['Max_HR']#交互特征df_all['Weight_Duration'] = df_all['Weight'] * df_all['Duration']df_all['Sex_Weight'] = df_all['Sex_encoded'] * df_all['Weight']# 构造运动功率特征df_all['workload'] = df_all['Weight'] * df_all['Duration'] * df_all['Heart_Rate'] / 1000# 构造生理特征交互项df_all['age_heart_ratio'] = df_all['Age'] / df_all['Heart_Rate']# 时间维度特征（如有时间戳）df_all['hour_of_day'] = df_all['Duration']/60/24# 组合特征numeric_cols = df_all.columnsfor i in range(len(numeric_cols)):feature_1 = numeric_cols[i]for j in range(i + 1, len(numeric_cols)):feature_2 = numeric_cols[j]df_all[f'{feature_1}_x_{feature_2}'] = df_all[feature_1] * df_all[feature_2]#数值归一化#scaler = RobustScaler()#df_all = scaler.fit_transform(df_all)# 分箱，把连续变成离散的，看你在哪一类now_col = ['Age', 'Height', 'Weight', 'Duration', 'Heart_Rate', 'Body_Temp', 'BMI']for i in now_col:df_all[i + "_box"] = pd.cut(df_all[i], bins=10, labels=False, right=False)baseline_temp = 37.0# Calculate 'Temp_Change' for the training datadf_all['Temp_Change'] = df_all['Body_Temp'] - baseline_temp# Calculate 'Intensity' for the training datadf_all['Intensity'] = df_all['Heart_Rate'] / df_all['Duration']# Calculate 'Heart_Rate_Ratio' for the training datadf_all['Heart_Rate_Ratio'] = df_all['Heart_Rate'] / df_all['Age']# Calculate 'Duration_x_HeartRate' for the training datadf_all['Duration_x_HeartRate'] = df_all['Duration'] * df_all['Heart_Rate']# Calculate 'Weight_x_Duration' for the training datadf_all['Weight_x_Duration'] = df_all['Weight'] * df_all['Duration']# Calculate 'Height_x_Duration' for the training datadf_all['Height_x_Duration'] = df_all['Height'] * df_all['Duration']# Calculate 'Weight_x_Height' for the training datadf_all['Weight_x_Height'] = df_all['Weight'] * df_all['Height']# Calculate 'Weight_x_Intensity' for the training datadf_all['Weight_x_Intensity'] = df_all['Weight'] * df_all['Intensity']# Calculate 'Height_x_Intensity' for the training datadf_all['Height_x_Intensity'] = df_all['Height'] * df_all['Intensity']X_train = df_all[:df_train.shape[0]]Y_train = np.log1p(df_train['Calories'])x_test = df_all[df_train.shape[0]:]#xgbmodel_xgb =estimator=XGBRegressor(random_state=42,n_estimators=8000,objective='reg:squarederror',eval_metric='rmse',device='cuda',learning_rate=0.05,max_depth=8,colsample_bytree=0.75,subsample=0.9,#reg_lambda = 1,#reg_alpha = 0.5,early_stopping_rounds=200,)#lgbmodel_lgb = lightgbm.LGBMRegressor(n_estimators=3000,  # 增加迭代次数配合早停learning_rate=0.03,  # 减小学习率num_leaves=15,  # 限制模型复杂度min_child_samples=25,  # 增加最小叶子样本数reg_alpha=0.1,  # L1正则化reg_lambda=0.1,  # L2正则化objective='regression_l1',  # 改用MAE损失early_stopping_rounds=200,eval_metric='RMSE',)#catmodel_cat = CatBoostRegressor(iterations=3500,learning_rate=0.02,depth=12,loss_function='RMSE',l2_leaf_reg=3,random_seed=42,eval_metric='RMSE',verbose=1000,task_type='GPU',early_stopping_rounds=200,)#k折交叉print("🔄Generating Out-of-Fold (OOF) predictions and Test predictions for Base Models...\n" + "-"*70 + "\n")# --- Prediction Storage ---# Arrays to store out-of-fold (OOF)add_pred_val_cat = np.zeros(len(X_train))add_pred_val_xgb = np.zeros(len(X_train))add_pred_val_lgb = np.zeros(len(X_train))# Arrays to store test predictions (accumulated across folds for averaging)add_pred_test_cat = np.zeros(len(x_test))add_pred_test_xgb = np.zeros(len(x_test))add_pred_test_lgb = np.zeros(len(x_test))kf = KFold(n_splits=5, shuffle=True, random_state=42)for fold, (train_index, val_index) in enumerate(kf.split(X_train, Y_train)):print(f"\n---Fold {fold + 1}/{kf.n_splits} ---")x_train, x_val = X_train.iloc[train_index], X_train.iloc[val_index]y_train, y_val = Y_train.iloc[train_index], Y_train.iloc[val_index]#Apply log1p transformation to target for trainingy_train_log1p = np.log1p(y_train)y_val_log1p = np.log1p(y_val)# --- CatBoost Training and Prediction ---print("  ➡ Training CatBoost...")model_cat.fit(x_train, y_train_log1p,eval_set=[(x_val, y_val_log1p)],verbose=0  # Set to 100 if you want to see progress)pred_val_cat = model_cat.predict(x_val)pred_test_cat = model_cat.predict(x_test)# --- XGBoost Training and Prediction ---print("  ➡ Training XGBoost...")model_xgb.fit(x_train, y_train_log1p,eval_set=[(x_val, y_val_log1p)],verbose=0  # Set to 100 if you want to see progress)pred_val_xgb = model_xgb.predict(x_val)pred_test_xgb = model_xgb.predict(x_test)# --- LGBM Training and Prediction ---print("  ➡ Training LGBM...")model_lgb.fit(x_train, y_train_log1p,eval_set=[(x_val, y_val_log1p)],)pred_val_lgb = model_lgb.predict(x_val)pred_test_lgb = model_lgb.predict(x_test)# --- Store OOF and Test Predictions (transformed back to original scale) ---add_pred_val_cat[val_index] = np.expm1(pred_val_cat)add_pred_val_xgb[val_index] = np.expm1(pred_val_xgb)add_pred_val_lgb[val_index] = np.expm1(pred_val_lgb)add_pred_test_cat += np.expm1(pred_test_cat) / kf.n_splitsadd_pred_test_xgb += np.expm1(pred_test_xgb) / kf.n_splitsadd_pred_test_lgb += np.expm1(pred_test_lgb) / kf.n_splits# Ensure all predictions are non-negativeadd_pred_val_cat[add_pred_val_cat < 0] = 0add_pred_val_xgb[add_pred_val_xgb < 0] = 0add_pred_val_lgb[add_pred_val_lgb < 0] = 0# Note: test predictions will also be non-negative after final prediction step# Calculate and print RMSLE for individual models on this foldprint(f"  CatBoost RMSLE (Fold {fold + 1}): {rmsle(y_val, add_pred_val_cat[val_index]):.4f}")print(f"  XGBoost RMSLE (Fold {fold + 1}): {rmsle(y_val, add_pred_val_xgb[val_index]):.4f}")print(f"  LGBM RMSLE (Fold {fold + 1}): {rmsle(y_val, add_pred_val_lgb[val_index]):.4f}")x_meta_train = pd.DataFrame({'cat_pred': add_pred_val_cat,'xgb_pred': add_pred_val_xgb,'lgbm_pred': add_pred_val_lgb,})y_meta_train = Y_trainx_meta_test = pd.DataFrame({'cat_pred': add_pred_test_cat,'xgb_pred': add_pred_test_xgb,'lgbm_pred': add_pred_test_lgb,})model_meta = Ridge(random_state=42)model_meta.fit(x_meta_train, y_meta_train)print(f" meta RMSLE :{rmsle(y_meta_train, model_meta.predict(x_meta_train)):.4f}")pred_all = np.expm1(model_meta.predict(x_meta_test))submission = pd.DataFrame({'id': df_test['id'],'Calories': pred_all})submission['Calories'] = np.clip(submission['Calories'], a_min=df_test['Duration'], a_max=20*df_test['Duration'])submission.to_csv('submission.csv', index=False)

本文来自互联网用户投稿，该文观点仅代表作者本人，不代表本站立场。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如若转载，请注明出处：http://www.pswp.cn/web/81854.shtml

如若内容造成侵权/违法违规/事实不符，请联系多彩编程网进行投诉反馈email:809451989@qq.com，一经查实，立即删除！

相关文章

第十二篇：MySQL 分布式架构演进与云原生数据库探索

第十二篇：MySQL 分布式架构演进与云原生数据库探索

本篇聚焦 MySQL 在互联网架构演进过程中的角色变化&#xff0c;探讨其从单体向分布式、再向云原生架构转型的关键技术路径与实践建议。 一、传统单体架构下的 MySQL 应用模式 在早期项目中&#xff0c;MySQL 多用于中小型应用&#xff1a; 单节点部署&#xff1b; 水平扩展难…
阅读更多...
JVM——回顾：JVM的起源、特性与系统构成

JVM——回顾：JVM的起源、特性与系统构成

引入 在当今数字化时代&#xff0c;Java语言及其运行环境Java虚拟机&#xff08;JVM&#xff09;在软件开发领域占据着举足轻重的地位。从大型企业级应用到各类移动应用&#xff0c;JVM凭借其独特的特性和强大的功能&#xff0c;为开发者提供了高效且稳定的运行环境。 JVM的起…
阅读更多...
大疆上云API+流媒体服务器部署实现直播功能

大疆上云API+流媒体服务器部署实现直播功能

根据官网文档上云API&#xff0c;先将官方提供的Demo部署起来&#xff0c;后端和前端服务环境搭建请参考官方文档。因为官方文档没有对直播这块的环境搭建进行说明&#xff0c;所以下面主要对直播功能环境搭建做一个记录&#xff0c;仅供参考&#xff0c;如有不足之处&#xff…
阅读更多...
计算机网络 HTTP篇常见面试题总结

计算机网络 HTTP篇常见面试题总结

HTTP各版本区别 HTTP 1.0 无状态、无连接&#xff1a;每次请求都需要建立新的 TCP&#xff0c;处理完后立即关闭&#xff0c;导致开销较大。队头阻塞&#xff1a;每个请求必须按照顺序依次处理&#xff0c;前面的请求未完成&#xff0c;后面的请求只能等待&#xff0c;减低了…
阅读更多...
目标检测：YOLO 模型详解

目标检测：YOLO 模型详解

目录 一、YOLO&#xff08;You Only Look Once&#xff09;模型讲解 YOLOv1 YOLOv2 (YOLO9000) YOLOv3 YOLOv4 YOLOv5 YOLOv6 YOLOv7 YOLOv8 YOLOv9 YOLOv10 YOLOv11 YOLOv12 其他变体&#xff1a;PP-YOLO 二、YOLO 模型的 Backbone&#xff1a;Focus 结构 三、…
阅读更多...
开源 FcDesigner 表单设计器组件事件详解

开源 FcDesigner 表单设计器组件事件详解

FcDesigner 是一款基于Vue的开源低代码可视化表单设计器工具&#xff0c;通过数据驱动表单渲染。可以通过拖拽的方式快速创建表单&#xff0c;提高开发者对表单的开发效率&#xff0c;节省开发者的时间。并广泛应用于在政务系统、OA系统、ERP系统、电商系统、流程管理等领域。 …
阅读更多...
关于 smali：2. 从 Java 到 Smali 的映射

关于 smali：2. 从 Java 到 Smali 的映射

一、对照 Java 代码与 Smali 代码差异 1.1 方法调用差异&#xff1a;Java vs Smali Java 方法分类&#xff1a; 方法类型Java 示例Smali 指令特点说明静态方法Utils.print("hi")invoke-static没有 this 指针实例方法obj.show()invoke-virtual有 this&#xff0c;虚…
阅读更多...
2025年05月29日Github流行趋势

2025年05月29日Github流行趋势

项目名称&#xff1a;agenticSeek 项目地址url&#xff1a;https://github.com/Fosowl/agenticSeek项目语言&#xff1a;Python历史star数&#xff1a;11898今日star数&#xff1a;2379项目维护者&#xff1a;Fosowl, steveh8758, klimentij, ganeshnikhil, apps/copilot-pull-…
阅读更多...
Dubbo高频面试题

Dubbo高频面试题

引言 作为分布式服务框架的标杆&#xff0c;Dubbo凭借其高性能RPC通信、灵活的服务治理能力和丰富的容错机制&#xff0c;成为Java技术栈中微服务领域的核心考点。本文系统梳理Dubbo高频面试核心知识点&#xff0c;涵盖容错策略、负载均衡、注册中心原理、服务上下线感知等关键…
阅读更多...
氮气吹扫电磁阀

氮气吹扫电磁阀

一、氮气吹扫电磁阀的概述 氮气吹扫电磁阀是一种重要的工业自动控制设备&#xff0c;用于对工业设备中出现的杂质和沉淀物进行清理&#xff0c;以保证生产线的畅通和生产效率的稳定。其作用是在需要吹扫清洗的工业设备中&#xff0c;通过控制气源的气压&#xff0c;打开电磁阀…
阅读更多...
网络安全的守护者：iVX 如何构建全方位防护体系

网络安全的守护者：iVX 如何构建全方位防护体系

一、安全技术的三大趋势 在数字化时代&#xff0c;网络安全面临着前所未有的挑战。随着企业级应用的普及&#xff0c;安全技术也在不断演进。目前&#xff0c;安全技术架构的发展呈现出三大趋势&#xff1a; 零信任架构的崛起&#xff1a;传统的网络安全依赖于边界防护&#…
阅读更多...
微软云如何申请使用

微软云如何申请使用

微软云&#xff08;Azure&#xff09;新手“开荒”指南&#xff1a;5步搞定账户&#xff0c;直达云端&#xff01; 还在为云计算的复杂门槛发愁吗&#xff1f;别担心&#xff01;当全球83%的企业都在加速“上云”&#xff0c;微软智能云Azure凭借其在全球34个区域、200服务的庞…
阅读更多...
magic-api配置Git插件教程

magic-api配置Git插件教程

一、配置gitee.com 1&#xff0c;生成rsa密钥&#xff0c;在你的电脑右键使用管理员身份运行&#xff08;命令提示符&#xff09;&#xff0c;执行下面命令 ssh-keygen -t rsa -b 2048 -m PEM一直按回车键&#xff0c;不需要输入内容 找到 你电脑中的~/.ssh/id_rsa.pub 文件…
阅读更多...
ojs导入显示空白页错误信息

ojs导入显示空白页错误信息

ojs技术支持：ojs.net.cn error: Uncaught Error: Call to a member function getData() on null in /var/www/html/ojs3/classes/search/ArticleSearchIndex.inc.php:38 Stack trace: #0 /var/www/html/ojs3/plugins/importexport/esci/filter/esciXmlArticleFilter.inc.php(…
阅读更多...
【ConvLSTM第一期】ConvLSTM原理

【ConvLSTM第一期】ConvLSTM原理

目录 &#x1f9e0; 一、ConvLSTM 原理详解1.1 背景1.2 ConvLSTM 的结构 参考 ConvLSTM&#xff08;Convolutional Long Short-Term Memory&#xff09;是一种结合了卷积神经网络&#xff08;CNN&#xff09;与循环神经网络&#xff08;RNN&#xff09;中 LSTM&#xff08;长短…
阅读更多...
4.8.1 利用Spark SQL实现词频统计

4.8.1 利用Spark SQL实现词频统计

在利用Spark SQL实现词频统计的实战中&#xff0c;首先需要准备单词文件并上传至HDFS。接着&#xff0c;可以通过交互式方法或创建Spark项目来实现词频统计。交互式方法包括读取文本文件生成数据集&#xff0c;扁平化映射得到新数据集&#xff0c;然后将数据集转成数据帧&#…
阅读更多...
Linux相关概念和易错知识点（41）（UDP、TCP报头结构）

Linux相关概念和易错知识点（41）（UDP、TCP报头结构）

目录 1.UDP&#xff08;1&#xff09;传输层&#xff08;2&#xff09;UDP报头&#xff08;3&#xff09;缓冲区和sk_buff①缓冲区②sk_buff 2.TCP&#xff08;1&#xff09;发送和接受缓冲区&#xff08;2&#xff09;报头结构①按序到达②可靠传输③流量控制④紧急指针 1.UDP…
阅读更多...
光谱相机在生态修复监测中的应用

光谱相机在生态修复监测中的应用

光谱相机通过多维光谱数据采集与智能分析技术&#xff0c;在生态修复监测中构建起‌“感知-评估-验证”‌的全周期管理体系&#xff0c;其核心应用方向如下&#xff1a; 一、土壤修复效能量化评估 ‌重金属污染动态监测‌ 通过短波红外&#xff08;1000-2500nm&#xff09;波…
阅读更多...
[网页五子棋]项目介绍以及websocket的消息推送(轮询操作)、报文格式和握手过程(建立连接过程)

[网页五子棋]项目介绍以及websocket的消息推送(轮询操作)、报文格式和握手过程(建立连接过程)

文章目录 项目背景核心技术创建项目WebSocket消息推送轮询操作 报文格式握手过程(建立连接过程) 项目背景 用户模块 用户的注册和登录管理用户的天梯分数&#xff0c;比赛场数&#xff0c;获胜场数等信息 匹配模块 依据用户的天梯积分&#xff0c;来实现匹配机制 对战模块 把两…
阅读更多...
时序模型介绍

时序模型介绍

一.整体介绍 1.单变量 vs 多变量时序数据 单变量就是只根据时间预测&#xff0c;多变量还要考虑用户 2.为什么不能用机器学习预测&#xff1a; a.时间不是影响标签的关键因素 b.时间与标签之间的联系过于弱/过于复杂&#xff0c;因此时序模型依赖于时间与时间的相关性来进行预…
阅读更多...
最新文章

Copyright @ 2022~2023