สอนปลูกต้นไม้ในภาษา R (ภาค 3): 5 ขั้นตอนการสร้าง Boosted Trees ด้วย xgboost package ในภาษา R — ตัวอย่างการทำนายระดับการกินน้ำมันของรถใน mpg dataset

# Install
install.packages("xgboost")

# Load
library(xgboost)

# Install ggplot2
install.packages("ggplot2")

# Load ggplot2
library(ggplot2)

# Load the dataset
data(mpg)

# Preview
head(mpg)

# A tibble: 6 × 11
  manufacturer model displ  year   cyl trans      drv     cty   hwy fl    class  
  <chr>        <chr> <dbl> <int> <int> <chr>      <chr> <int> <int> <chr> <chr>  
1 audi         a4      1.8  1999     4 auto(l5)   f        18    29 p     compact
2 audi         a4      1.8  1999     4 manual(m5) f        21    29 p     compact
3 audi         a4      2    2008     4 manual(m6) f        20    31 p     compact
4 audi         a4      2    2008     4 auto(av)   f        21    30 p     compact
5 audi         a4      2.8  1999     6 auto(l5)   f        16    26 p     compact
6 audi         a4      2.8  1999     6 manual(m5) f        18    26 p     compact

# View the tructure
str(mpg)

tibble [234 × 11] (S3: tbl_df/tbl/data.frame)
 $ manufacturer: chr [1:234] "audi" "audi" "audi" "audi" ...
 $ model       : chr [1:234] "a4" "a4" "a4" "a4" ...
 $ displ       : num [1:234] 1.8 1.8 2 2 2.8 2.8 3.1 1.8 1.8 2 ...
 $ year        : int [1:234] 1999 1999 2008 2008 1999 1999 2008 1999 1999 2008 ...
 $ cyl         : int [1:234] 4 4 4 4 6 6 6 4 4 4 ...
 $ trans       : chr [1:234] "auto(l5)" "manual(m5)" "manual(m6)" "auto(av)" ...
 $ drv         : chr [1:234] "f" "f" "f" "f" ...
 $ cty         : int [1:234] 18 21 20 21 16 18 18 18 16 20 ...
 $ hwy         : int [1:234] 29 29 31 30 26 26 27 26 25 28 ...
 $ fl          : chr [1:234] "p" "p" "p" "p" ...
 $ class       : chr [1:234] "compact" "compact" "compact" "compact" ...

# Convert character columns to factor

## Get character columns
chr_cols <- c("manufacturer",
              "model",
              "trans",
              "drv",
              "fl",
              "class")

## For-loop through the character columns
for (col in chr_cols) {
  mpg[[col]] <- as.factor(mpg[[col]])
}

## Check the results
str(mpg)

tibble [234 × 11] (S3: tbl_df/tbl/data.frame)
 $ manufacturer: Factor w/ 15 levels "audi","chevrolet",..: 1 1 1 1 1 1 1 1 1 1 ...
 $ model       : Factor w/ 38 levels "4runner 4wd",..: 2 2 2 2 2 2 2 3 3 3 ...
 $ displ       : num [1:234] 1.8 1.8 2 2 2.8 2.8 3.1 1.8 1.8 2 ...
 $ year        : int [1:234] 1999 1999 2008 2008 1999 1999 2008 1999 1999 2008 ...
 $ cyl         : int [1:234] 4 4 4 4 6 6 6 4 4 4 ...
 $ trans       : Factor w/ 10 levels "auto(av)","auto(l3)",..: 4 9 10 1 4 9 1 9 4 10 ...
 $ drv         : Factor w/ 3 levels "4","f","r": 2 2 2 2 2 2 2 1 1 1 ...
 $ cty         : int [1:234] 18 21 20 21 16 18 18 18 16 20 ...
 $ hwy         : int [1:234] 29 29 31 30 26 26 27 26 25 28 ...
 $ fl          : Factor w/ 5 levels "c","d","e","p",..: 4 4 4 4 4 4 4 4 4 4 ...
 $ class       : Factor w/ 7 levels "2seater","compact",..: 2 2 2 2 2 2 2 2 2 2 ...

# Separate the features from the outcome

## Get the features
x <- mpg[, !names(mpg) %in% "hwy"]

## One-hot encode the features
x <- model.matrix(~ . - 1,
                  data = x)

## Get the outcome
y <- mpg$hwy

# Split the data

## Set seed for reproducibility
set.seed(360)

## Get training index
train_index <- sample(1:nrow(x),
                      nrow(x) * 0.8)

## Create x, y train
x_train <- x[train_index, ]
y_train <- y[train_index]

## Create x, y test
x_test <- x[-train_index, ]
y_test <- y[-train_index]

## Check the results
cat("TRAIN SET", "\\n")
cat("1. Data in x_train:", nrow(x_train), "\\n")
cat("2. Data in y_train:", length(y_train), "\\n")
cat("---", "\\n", "TEST SET", "\\n")
cat("1. Data in x_test:", nrow(x_test), "\\n")
cat("2. Data in y_test:", length(y_test), "\\n")

TRAIN SET
1. Data in x_train: 187
2. Data in y_train: 187
---
TEST SET
1. Data in x_test: 47
2. Data in y_test: 47

# Convert to DMatrix

## Training set
train_set <- xgb.DMatrix(data = x_train,
                         label = y_train)

## Test set
test_set <- xgb.DMatrix(data = x_test,
                        label = y_test)

## Check the results
train_set
test_set

TRAIN SET
xgb.DMatrix  dim: 187 x 77  info: label  colnames: yes
---
TEST SET
xgb.DMatrix  dim: 47 x 77  info: label  colnames: yes

# Train the model

## Set hyperparametres
hp <- list(objective = "reg:squarederror",
           eta = 0.1,
           max_depth = 4,
           eval_metric = c("rmse",
                           "mae"))

## Train
xgb_model <- xgb.train(params = hp,
                       data = train_set,
                       nrounds = 50,
                       watchlist = list(train = train_set,
                                        test = test_set),
                       verbose = 1)

[1]	train-rmse:21.083975	test-rmse:22.739357 
[2]	train-rmse:19.045063	test-rmse:20.598582 
[3]	train-rmse:17.204130	test-rmse:18.713079 
[4]	train-rmse:15.549113	test-rmse:16.974701 
[5]	train-rmse:14.053049	test-rmse:15.453560 
[6]	train-rmse:12.707307	test-rmse:14.097377 
[7]	train-rmse:11.495216	test-rmse:12.877722 
[8]	train-rmse:10.402476	test-rmse:11.767320 
[9]	train-rmse:9.413522	test-rmse:10.740546 
[10]	train-rmse:8.525230	test-rmse:9.863130 
[11]	train-rmse:7.722776	test-rmse:9.068840 
[12]	train-rmse:7.000648	test-rmse:8.357181 
[13]	train-rmse:6.346603	test-rmse:7.687483 
[14]	train-rmse:5.758685	test-rmse:7.091249 
[15]	train-rmse:5.229548	test-rmse:6.557082 
[16]	train-rmse:4.753713	test-rmse:6.079389 
[17]	train-rmse:4.325653	test-rmse:5.651858 
[18]	train-rmse:3.940325	test-rmse:5.275154 
[19]	train-rmse:3.594545	test-rmse:4.938849 
[20]	train-rmse:3.283961	test-rmse:4.627743 
[21]	train-rmse:3.003089	test-rmse:4.352060 
[22]	train-rmse:2.747553	test-rmse:4.110172 
[23]	train-rmse:2.519617	test-rmse:3.889650 
[24]	train-rmse:2.314957	test-rmse:3.691806 
[25]	train-rmse:2.133630	test-rmse:3.499208 
[26]	train-rmse:1.969083	test-rmse:3.330280 
[27]	train-rmse:1.823011	test-rmse:3.181541 
[28]	train-rmse:1.693565	test-rmse:3.045308 
[29]	train-rmse:1.575817	test-rmse:2.919070 
[30]	train-rmse:1.469256	test-rmse:2.812063 
[31]	train-rmse:1.375599	test-rmse:2.700515 
[32]	train-rmse:1.292928	test-rmse:2.615973 
[33]	train-rmse:1.218867	test-rmse:2.541929 
[34]	train-rmse:1.151134	test-rmse:2.462113 
[35]	train-rmse:1.092395	test-rmse:2.404873 
[36]	train-rmse:1.039158	test-rmse:2.336600 
[37]	train-rmse:0.993882	test-rmse:2.291398 
[38]	train-rmse:0.952062	test-rmse:2.236936 
[39]	train-rmse:0.915935	test-rmse:2.198657 
[40]	train-rmse:0.879957	test-rmse:2.152984 
[41]	train-rmse:0.850423	test-rmse:2.102272 
[42]	train-rmse:0.822475	test-rmse:2.054172 
[43]	train-rmse:0.799025	test-rmse:2.011621 
[44]	train-rmse:0.775398	test-rmse:1.971787 
[45]	train-rmse:0.755066	test-rmse:1.933539 
[46]	train-rmse:0.736655	test-rmse:1.900084 
[47]	train-rmse:0.719087	test-rmse:1.870832 
[48]	train-rmse:0.705279	test-rmse:1.853400 
[49]	train-rmse:0.691914	test-rmse:1.834918 
[50]	train-rmse:0.680016	test-rmse:1.825738 

# Print the model
xgb_model

##### xgb.Booster
raw: 62.4 Kb 
call:
  xgb.train(params = hp, data = train_set, nrounds = 50, watchlist = list(train = train_set, 
    test = test_set), verbose = 1)
params (as set within xgb.train):
  objective = "reg:squarederror", eta = "0.1", max_depth = "4", eval_metric = "rmse", validate_parameters = "mae", objective = "TRUE"
xgb.attributes:
  niter
callbacks:
  cb.print.evaluation(period = print_every_n)
  cb.evaluation.log()
# of features: 77 
niter: 50
nfeatures : 77 
evaluation_log:
  iter train_rmse test_rmse
 <num>      <num>     <num>
     1 21.0839746 22.739357
     2 19.0450628 20.598582
   ---        ---       ---
    49  0.6919137  1.834918
    50  0.6800159  1.825738

# Make predictions
y_pred <- predict(xgb_model,
                  newdata = x_test)

# Compare predictions to actual outcomes
results <- data.frame(actual = y_test,
                      predicted = y_pred,
                      error = y_test - y_pred)

# Preview the results
head(results, 10)

   actual predicted      error
1      31  27.81219  3.1878090
2      25  25.89449 -0.8944893
3      30  30.13318 -0.1331844
4      29  26.77814  2.2218552
5      24  24.34723 -0.3472347
6      23  23.58175 -0.5817528
7      19  17.81131  1.1886921
8      12  12.32908 -0.3290768
9      12  12.31534 -0.3153391
10     16  16.25793 -0.2579288

# Calculate MAE
mae <- mean(abs(results$error))

# Calculate RMSE
rmse <- sqrt(mean((results$error)^2))

# Calculate R squared
ss_res <- sum((results$error)^2)
ss_tot <- sum((results$actual - mean(results$actual))^2)
r_squared <- 1 - (ss_res / ss_tot)

# Print the results
cat("MAE:", round(mae, 2), "\\n")
cat("RMSE:", round(rmse, 2), "\\n")
cat("R squared:", round(r_squared, 2), "\\n")

MAE: 1.23
RMSE: 1.83
R squared: 0.93