很多人戏称调参的过程就像是"炼丹"！确实差不多，而且很多时候你调整后的结果可能还不如默认的结果好！这就好比打游戏，"一顿操作猛如虎，一看战绩0比5"！
模型调优一定要基于对算法和数据的理解进行，不是随便调的。
我们使用著名的糖尿病数据集进行演示，首先创建任务
library(mlr3verse)
## 载入需要的程辑包：mlr3
task <- tsk("pima")
print(task)
## (768 x 9)
## * Target: diabetes
## * Properties: twoclass
## * Features (8):
## - dbl (8): age, glucose, insulin, mass, pedigree, pregnant, pressure,
## triceps
选择算法,查看算法支持的超参数
learner <- lrn("classif.rpart")
learner$param_set
##
## id class lower upper nlevels default value
## 1: cp ParamDbl 0 1 Inf 0.01
## 2: keep_model ParamLgl NA NA 2 FALSE
## 3: maxcompete ParamInt 0 Inf Inf 4
## 4: maxdepth ParamInt 1 30 30 30
## 5: maxsurrogate ParamInt 0 Inf Inf 5
## 6: minbucket ParamInt 1 Inf Inf
## 7: minsplit ParamInt 1 Inf Inf 20
## 8: surrogatestyle ParamInt 0 1 2 0
## 9: usesurrogate ParamInt 0 2 3 2
## 10: xval ParamInt 0 Inf Inf 10 0
在这里我们选择调整复杂度参数cp和最小分支参数minsplit，并设定超参数的调整范围：
search_space <- ps(
cp = p_dbl(lower = 0.001, upper = 0.1),
minsplit = p_int(lower = 1, upper = 10)
)
search_space
##
## id class lower upper nlevels default value
## 1: cp ParamDbl 0.001 0.1 Inf
## 2: minsplit ParamInt 1.000 10.0 10
然后选择重抽样方法和性能指标
hout <- rsmp("holdout", ratio = 0.7)
measure <- msr("classif.ce")
接下来进行调参有两种方法。
方法一：通过tuninginstancesinglecrite和tuner训练模型
library(mlr3tuning)
## 载入需要的程辑包：paradox
evals20 <- trm("evals", n_evals = 20) # 设定何时停止训练
# 统一放入instance中
instance <- TuningInstanceSingleCrit$new(
task = task,
learner = learner,
resampling = hout,
measure = measure,
terminator = evals20,
search_space = search_space
)
instance
##
## * State: Not optimized
## * Objective:
## * Search Space:
##
## id class lower upper nlevels default value
## 1: cp ParamDbl 0.001 0.1 Inf
## 2: minsplit ParamInt 1.000 10.0 10
## * Terminator:
## * Terminated: FALSE
## * Archive:
##
## Null data.table (0 rows and 0 cols)
关于何时停止训练，mlr3给出了5种方法：
- Terminate after a given time：一定时间后停止
- Terninate after a given number of iterations：特定迭代次数后停止
- Terminate after a specific performance has been reached：达到特定性能指标后停止
- Terminate when tuning dose find a better configuration for a given number of iterations：在给定迭代次数中确实找到表现很好的参数组合后停止
- A combination of above in ALL or ANY fashon：上面几种方法组合
然后还需要设置超参数搜索的方法：
mlr3tuning目前支持以下超参数搜索的方法：
- Grid search：网格搜索
- Random search：随机搜索
- Generalized simulated annealing
- Non-Linear optimization
# 这里选择网格搜索
tuner <- tnr("grid_search", resolution = 5) # 网格搜索
接下来就是进行训练模型，上面我们设置了网格搜索的分辨率是5，我们有2个超参数需要调整，所以理论上一共有5 * 5 = 25个组合，但是在前面的停止搜索的方法中我们选择了n_evals = 20，所有实际上在评价完20个组合后就会停止了！
#lgr::get_logger("mlr3")$set_threshold("warn")
#lgr::get_logger("bbotk")$set_threshold("warn") # 减少屏幕打印内容
tuner$optimize(instance)
## INFO [20:51:28.312] [bbotk] Starting to optimize 2 parameter(s) with '' and ' [n_evals=20, k=0]'
## INFO [20:51:28.331] [bbotk] Evaluating 1 configuration(s)
## INFO [20:51:29.309] [bbotk] Finished optimizing after 20 evaluation(s)
## INFO [20:51:29.310] [bbotk] Result:
## INFO [20:51:29.310] [bbotk] cp minsplit learner_param_vals x_domain classif.ce
## INFO [20:51:29.310] [bbotk] 0.02575 3 0.2130435
## cp minsplit learner_param_vals x_domain classif.ce
## 1: 0.02575 3 0.2130435
查看调整好的超参数：
instance$result_learner_param_vals
## $xval
## [1] 0
##
## $cp
## [1] 0.02575
##
## $minsplit
## [1] 3
查看模型性能：
instance$result_y
## classif.ce
## 0.2130435
查看每一次迭代的结果，只有20个：
instance$archive
##
## cp minsplit classif.ce runtime_learners timestamp batch_nr
## 1: 0.026 3 0.21 0.02 2022-02-27 20:51:28 1
## 2: 0.075 8 0.21 0.00 2022-02-27 20:51:28 2
## 3: 0.050 5 0.21 0.00 2022-02-27 20:51:28 3
## 4: 0.001 1 0.30 0.00 2022-02-27 20:51:28 4
## 5: 0.100 3 0.21 0.02 2022-02-27 20:51:28 5
## 6: 0.026 5 0.21 0.02 2022-02-27 20:51:28 6
## 7: 0.100 8 0.21 0.01 2022-02-27 20:51:28 7
## 8: 0.001 8 0.27 0.00 2022-02-27 20:51:28 8
## 9: 0.001 5 0.28 0.00 2022-02-27 20:51:28 9
## 10: 0.100 5 0.21 0.02 2022-02-27 20:51:28 10
## 11: 0.075 10 0.21 0.00 2022-02-27 20:51:28 11
## 12: 0.050 10 0.21 0.01 2022-02-27 20:51:28 12
## 13: 0.075 5 0.21 0.00 2022-02-27 20:51:28 13
## 14: 0.050 8 0.21 0.01 2022-02-27 20:51:29 14
## 15: 0.001 10 0.26 0.00 2022-02-27 20:51:29 15
## 16: 0.050 3 0.21 0.00 2022-02-27 20:51:29 16
## 17: 0.050 1 0.21 0.02 2022-02-27 20:51:29 17
## 18: 0.100 10 0.21 0.00 2022-02-27 20:51:29 18
## 19: 0.075 1 0.21 0.01 2022-02-27 20:51:29 19
## 20: 0.026 1 0.21 0.00 2022-02-27 20:51:29 20
## warnings errors resample_result
## 1: 0 0
## 2: 0 0
## 3: 0 0
## 4: 0 0
## 5: 0 0
## 6: 0 0
## 7: 0 0
## 8: 0 0
## 9: 0 0
## 10: 0 0
## 11: 0 0
## 12: 0 0
## 13: 0 0
## 14: 0 0
## 15: 0 0
## 16: 0 0
## 17: 0 0
## 18: 0 0
## 19: 0 0
## 20: 0 0
接下来就可以把训练好的超参数应用于模型，重新应用于数据：
learner$param_set$values <- instance$result_learner_param_vals
learner$train(task)
这个训练好的模型就可以用于预测了，使用learner$predict()即可！
以上步骤写起来有些复杂，与tidymodels相比不够简洁好理解，我刚开始学习的时候经常记不住，后来版本更新后终于有了简便写法：
instance <- tune(
task = task,
learner = learner,
resampling = hout,
measure = measure,
search_space = search_space,
method = "grid_search",
resolution = 5,
term_evals = 25
)
## INFO [20:51:29.402] [bbotk] Starting to optimize 2 parameter(s) with '' and ' [n_evals=25, k=0]'
## INFO [20:51:29.403] [bbotk] Evaluating 1 configuration(s)
## INFO [20:51:30.534] [bbotk] Finished optimizing after 25 evaluation(s)
## INFO [20:51:30.534] [bbotk] Result:
## INFO [20:51:30.535] [bbotk] cp minsplit learner_param_vals x_domain classif.ce
## INFO [20:51:30.535] [bbotk] 0.02575 10 0.2347826
instance$result_learner_param_vals
## $xval
## [1] 0
##
## $cp
## [1] 0.02575
##
## $minsplit
## [1] 10
instance$result_y
## classif.ce
## 0.2347826
learner$param_set$values <- instance$result_learner_param_vals
learner$train(task)
mlr3也支持同时设定多个性能指标：
measures <- msrs(c("classif.ce","time_train")) # 设定多个评价指标
evals20 <- trm("evals", n_evals = 20)
instance <- TuningInstanceMultiCrit$new(
task = task,
learner = learner,
resampling = hout,
measures = measures,
search_space = search_space,
terminator = evals20
)
tuner$optimize(instance)
## INFO [20:51:30.595] [bbotk] Starting to optimize 2 parameter(s) with '' and ' [n_evals=20, k=0]'
## INFO [20:51:30.597] [bbotk] Evaluating 1 configuration(s)
## INFO [20:51:30.605] [mlr3] Running benchmark with 1 resampling iterations
## INFO [20:51:30.608] [mlr3] Applying learner 'classif.rpart' on task 'pima' (iter 1/1)
## INFO [20:51:30.620] [mlr3] Finished benchmark
## INFO [20:51:30.642] [bbotk] Result of batch 1:
## INFO [20:51:30.643] [bbotk] cp minsplit classif.ce time_train warnings errors runtime_learners
## INFO [20:51:30.643] [bbotk] 0.0505 1 0.2347826 0 0 0 0.02
## cp minsplit learner_param_vals x_domain classif.ce time_train
## 1: 0.05050 1 0.2347826 0
## 2: 0.07525 1 0.2347826 0
## 3: 0.07525 10 0.2347826 0
## 4: 0.10000 8 0.2347826 0
## 5: 0.02575 3 0.2347826 0
## 6: 0.07525 8 0.2347826 0
## 7: 0.10000 3 0.2347826 0
## 8: 0.10000 5 0.2347826 0
## 9: 0.02575 5 0.2347826 0
## 10: 0.07525 5 0.2347826 0
## 11: 0.05050 8 0.2347826 0
## 12: 0.05050 3 0.2347826 0
## 13: 0.07525 3 0.2347826 0
## 14: 0.05050 5 0.2347826 0
## 15: 0.02575 1 0.2347826 0
查看结果：
instance$result_learner_param_vals
## [[1]]
## [[1]]$xval
## [1] 0
##
## [[1]]$cp
## [1] 0.0505
##
## [[1]]$minsplit
## [1] 1
##
##
## [[2]]
## [[2]]$xval
## [1] 0
##
## [[2]]$cp
## [1] 0.07525
##
## [[2]]$minsplit
## [1] 1
##
##
## [[3]]
## [[3]]$xval
## [1] 0
##
## [[3]]$cp
## [1] 0.07525
##
## [[3]]$minsplit
## [1] 10
##
##
## [[4]]
## [[4]]$xval
## [1] 0
##
## [[4]]$cp
## [1] 0.1
##
## [[4]]$minsplit
## [1] 8
##
##
## [[5]]
## [[5]]$xval
## [1] 0
##
## [[5]]$cp
## [1] 0.02575
##
## [[5]]$minsplit
## [1] 3
##
##
## [[6]]
## [[6]]$xval
## [1] 0
##
## [[6]]$cp
## [1] 0.07525
##
## [[6]]$minsplit
## [1] 8
##
##
## [[7]]
## [[7]]$xval
## [1] 0
##
## [[7]]$cp
## [1] 0.1
##
## [[7]]$minsplit
## [1] 3
##
##
## [[8]]
## [[8]]$xval
## [1] 0
##
## [[8]]$cp
## [1] 0.1
##
## [[8]]$minsplit
## [1] 5
##
##
## [[9]]
## [[9]]$xval
## [1] 0
##
## [[9]]$cp
## [1] 0.02575
##
## [[9]]$minsplit
## [1] 5
##
##
## [[10]]
## [[10]]$xval
## [1] 0
##
## [[10]]$cp
## [1] 0.07525
##
## [[10]]$minsplit
## [1] 5
##
##
## [[11]]
## [[11]]$xval
## [1] 0
##
## [[11]]$cp
## [1] 0.0505
##
## [[11]]$minsplit
## [1] 8
##
##
## [[12]]
## [[12]]$xval
## [1] 0
##
## [[12]]$cp
## [1] 0.0505
##
## [[12]]$minsplit
## [1] 3
##
##
## [[13]]
## [[13]]$xval
## [1] 0
##
## [[13]]$cp
## [1] 0.07525
##
## [[13]]$minsplit
## [1] 3
##
##
## [[14]]
## [[14]]$xval
## [1] 0
##
## [[14]]$cp
## [1] 0.0505
##
## [[14]]$minsplit
## [1] 5
##
##
## [[15]]
## [[15]]$xval
## [1] 0
##
## [[15]]$cp
## [1] 0.02575
##
## [[15]]$minsplit
## [1] 1
instance$rusult_y
## NULL
以上就是第一种方法，接下来介绍第二种方法。
方法二：通过autotuner训练模型
这种方式方法把调整参数、将调整好的参数应用于模型放到一起了，但是也需要提前设定好各种需要的参数。
task <- tsk("pima") # 创建任务
leanrer <- lrn("classif.rpart") # 选择学习器
search_space <- ps(
cp = p_dbl(0.001, 0.1),
minsplit = p_int(1,10)
) # 设定搜索范围
terminator <- trm("evals", n_evals = 10) # 设定停止标志
tuner <- tnr("random_search") # 选择搜索方法
resampling <- rsmp("holdout") # 选择重抽样方法
measure <- msr("classif.acc") # 选择评价指标
# 训练
at <- AutoTuner$new(
learner = learner,
resampling = resampling,
search_space = search_space,
measure = measure,
tuner = tuner,
terminator = terminator
)
自动选择最优参数并作用于数据：
at$train(task)
## INFO [20:51:31.873] [bbotk] Starting to optimize 2 parameter(s) with '' and ' [n_evals=10, k=0]'
## INFO [20:51:32.332] [bbotk] 0.02278977 3 0.7695312
at$predict(task)
## for 768 observations:
## row_ids truth response
## 1 pos pos
## 2 neg neg
## 3 pos neg
## ---
## 766 neg neg
## 767 pos neg
## 768 neg neg
这个方法也有个简便写法：
auto_learner <- auto_tuner(
learner = learner,
resampling = resampling,
measure = measure,
search_space = search_space,
method = "random_search",
term_evals = 10
)
auto_learner$train(task)
## INFO [20:51:32.407] [bbotk] Starting to optimize 2 parameter(s) with '' and ' [n_evals=10, k=0]'
## INFO [20:51:32.858] [bbotk] Finished optimizing after 10 evaluation(s)
## INFO [20:51:32.859] [bbotk] Result:
## INFO [20:51:32.859] [bbotk] cp minsplit learner_param_vals x_domain classif.acc
## INFO [20:51:32.859] [bbotk] 0.02922122 8 0.7539062
auto_learner$predict(task)
## for 768 observations:
## row_ids truth response
## 1 pos pos
## 2 neg neg
## 3 pos neg
## ---
## 766 neg neg
## 767 pos neg
## 768 neg neg
超参数设定的方法
每次单独设置超参数的范围等可能会显得比较笨重无聊，mlr3也提供另外一种可以在选择学习器时进行设定超参数的方法。
# 在选择学习器时设置超参数范围
learner <- lrn("classif.svm")
learner$param_set$values$kernel <- "polynomial"
learner$param_set$values$degree <- to_tune(lower = 1, upper = 3)
print(learner$param_set$search_space())
##
## id class lower upper nlevels default value
## 1: degree ParamInt 1 3 3
但其实这样也有问题，这个方法要求你对算法很熟悉，能够记住所有超参数记忆它们在mlr3中的拼写！但很显然这有点困难，所有我还是推荐第一种，每次单独设置，记不住还可以查看一下具体的超参数。
参数依赖
某些超参数只有在某些条件下才有效，比如支持向量机（SVM），它的degree参数只有在kernel是polynomial时才有效，这种情况也可以在mlr3中设置好。
library(data.table)
search_space = ps(
cost = p_dbl(-1, 1, trafo = function(x) 10^x), # 可进行数据变换
kernel = p_fct(c("polynomial", "radial")),
degree = p_int(1, 3, depends = kernel == "polynomial") # 设置参数依赖
)
rbindlist(generate_design_grid(search_space, 3)$transpose(), fill = TRUE)
## cost kernel degree
## 1: 0.1 polynomial 1
## 2: 0.1 polynomial 2
## 3: 0.1 polynomial 3
## 4: 0.1 radial NA
## 5: 1.0 polynomial 1
## 6: 1.0 polynomial 2
## 7: 1.0 polynomial 3
## 8: 1.0 radial NA
## 9: 10.0 polynomial 1
## 10: 10.0 polynomial 2
## 11: 10.0 polynomial 3
## 12: 10.0 radial NA
进行以上设置后在进行后面的操作时不会出错，自动处理。