R语言|快速构建机器学习模型、评估与比较（fastml）

# 加载R包library(fastml)

fastml(  data, # 数据集  label, # 目标变量  algorithms = c("xgboost", "random_forest", "svm_radial"), # 支持的算法  test_size = 0.2, # 测试集的比例  resampling_method = "cv", # 重采样方法  folds = 5, # 交叉验证的折数  tune_params = NULL, # 调参  metric = "Accuracy", # 评估指标  n_cores = 1,  # CPU核数  stratify = TRUE,  # 分层抽样  impute_method = NULL, # 缺失数据插补方法  encode_categoricals = TRUE, # 分类变量是否编码  scaling_methods = c("center", "scale"), # 数据归一化  summaryFunction = NULL,  # 自定义数据摘要方法  seed = 123  # 种子数)

data(iris) # 载入数据
iris$Species <- factor(iris$Species) # 转换为因子

# 模型训练model <- fastml(  data = iris,   #数据集  label = "Species",  # 目标变量  n_cores = 6  )

# 模型训练结果summary(model)

===== fastml Model Summary =====Best Model: random_forest 
Performance Metrics for All Models:
         Model Accuracy Kappa Sensitivity Specificity Precision random_forest     0.95   0.9         0.9           1         1       xgboost     0.90   0.8         0.8           1         1    svm_radial     0.90   0.8         0.8           1         1        F1 0.9473684 0.8888889 0.8888889
Best Model Hyperparameters:
 mtry    2
Generating performance comparison plots...
To make predictions, use the 'predict' function.=================================

library(modeldata)data("credit_data")
# 模型训练model <- fastml(  data = na.omit(credit_data),  label = "Status",  algorithms = "xgboost",   # 指定算法  resampling_method = "cv",  folds = 3,  n_cores = 8,  test_size = 0.25)
# View model summarysummary(model)

===== fastml Model Summary =====Best Model: xgboost 
Performance Metrics for All Models:
   Model  Accuracy     Kappa Sensitivity Specificity Precision        F1 xgboost 0.7998018 0.4002036   0.4257812   0.9269588 0.6646341 0.5190476
Best Model Hyperparameters:
 nrounds max_depth eta gamma colsample_bytree min_child_weight subsample     100         5 0.1     0              0.6                5       0.6
Generating performance comparison plots...
To make predictions, use the 'predict' function.=================================

# 自定义超参数搜索custom_tuning <- list(  random_forest = expand.grid(mtry = c(1:10)))
model <- fastml(  data = iris,  label = "Species",  algorithms = c("random_forest"),  tune_params = custom_tuning,  metric = "Accuracy")
# View the resultssummary(model)
===== fastml Model Summary =====Best Model: random_forest 
Performance Metrics for All Models:
         Model Accuracy Kappa Sensitivity Specificity Precision random_forest      0.9   0.8         0.8           1         1        F1 0.8888889
Best Model Hyperparameters:
 mtry    1
Generating performance comparison plots...
To make predictions, use the 'predict' function.=================================

predict(model,iris)

[1] versicolor versicolor versicolor versicolor versicolor  [6] versicolor versicolor versicolor versicolor versicolor [11] versicolor versicolor versicolor versicolor versicolor [16] versicolor versicolor versicolor versicolor versicolor [21] versicolor versicolor virginica  versicolor versicolor [26] versicolor versicolor versicolor versicolor versicolor [31] versicolor versicolor versicolor virginica  versicolor [36] versicolor versicolor versicolor versicolor versicolor [41] versicolor versicolor versicolor versicolor versicolor [46] versicolor versicolor versicolor versicolor versicolor [51] virginica  virginica  virginica  virginica  virginica  [56] virginica  virginica  virginica  virginica  virginica  [61] virginica  virginica  virginica  virginica  virginica  [66] virginica  virginica  virginica  virginica  virginica  [71] virginica  virginica  virginica  virginica  virginica  [76] virginica  virginica  virginica  virginica  virginica  [81] virginica  virginica  virginica  virginica  virginica  [86] virginica  virginica  virginica  virginica  virginica  [91] virginica  virginica  virginica  virginica  virginica  [96] virginica  virginica  virginica  virginica  virginica Levels: versicolor virginica