Modul 4 Analisis Diskriminan dan Regresi Logistik Multinomial pada Dataset Balance Scale
Ananta Putri Clara (24031554031), Ayu Anggraini (24031554054), Faizah Khasifa Salsabila (24031554056)
LOAD DATA & PREPROCESSING
df <- read.csv("balance-scale.data", header = FALSE)Data dibaca dari file balance-scale.data menggunakan fungsi read.csv() tanpa header untuk digunakan dalam analisis selanjutnya.
colnames(df) <- c("class", "left_weight", "left_distance", "right_weight", "right_distance")Memberikan nama kolom pada dataset agar lebih mudah dipahami.
df$class <- as.factor(df$class)Mengubah variabel class menjadi tipe faktor untuk keperluan klasifikasi.
str(df)## 'data.frame': 625 obs. of 5 variables:
## $ class : Factor w/ 3 levels "B","L","R": 1 3 3 3 3 3 3 3 3 3 ...
## $ left_weight : int 1 1 1 1 1 1 1 1 1 1 ...
## $ left_distance : int 1 1 1 1 1 1 1 1 1 1 ...
## $ right_weight : int 1 1 1 1 1 2 2 2 2 2 ...
## $ right_distance: int 1 2 3 4 5 1 2 3 4 5 ...Menampilkan struktur data untuk melihat tipe dan isi setiap variabel.
table(df$class)##
## B L R
## 49 288 288Menampilkan jumlah data pada masing-masing kelas.
summary(df)## class left_weight left_distance right_weight right_distance
## B: 49 Min. :1 Min. :1 Min. :1 Min. :1
## L:288 1st Qu.:2 1st Qu.:2 1st Qu.:2 1st Qu.:2
## R:288 Median :3 Median :3 Median :3 Median :3
## Mean :3 Mean :3 Mean :3 Mean :3
## 3rd Qu.:4 3rd Qu.:4 3rd Qu.:4 3rd Qu.:4
## Max. :5 Max. :5 Max. :5 Max. :5Menampilkan ringkasan statistik deskriptif dari setiap variabel.
SPLIT DATA
set.seed(123)
index <- sample(1:nrow(df), 0.7 * nrow(df))Membuat indeks acak untuk membagi data menjadi data latih (70%).
train <- df[index, ]
test <- df[-index, ]Membagi data menjadi data latih dan data uji.
UJI ASUMSI
install.packages("MVN")## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.5'
## (as 'lib' is unspecified)Menginstal package MVN untuk uji normalitas multivariat.
library(MVN)## Registered S3 method overwritten by 'lme4':
## method from
## na.action.merMod carMemanggil package MVN.
mvn_result <- MVN::mvn(
data = train[, -1],
mvn_test = "mardia"
)Melakukan uji normalitas multivariat menggunakan metode Mardia.
mvn_result## $multivariate_normality
## Test Statistic p.value Method MVN
## 1 Mardia Skewness 2.125 1 asymptotic ✓ Normal
## 2 Mardia Kurtosis -7.898 <0.001 asymptotic ✗ Not normal
##
## $univariate_normality
## Test Variable Statistic p.value Normality
## 1 Anderson-Darling left_weight 15.994 <0.001 ✗ Not normal
## 2 Anderson-Darling left_distance 15.606 <0.001 ✗ Not normal
## 3 Anderson-Darling right_weight 15.249 <0.001 ✗ Not normal
## 4 Anderson-Darling right_distance 16.562 <0.001 ✗ Not normal
##
## $descriptives
## Variable n Mean Std.Dev Median Min Max 25th 75th Skew Kurtosis
## 1 left_weight 437 2.938 1.420 3 1 5 2 4 0.071 1.692
## 2 left_distance 437 3.032 1.406 3 1 5 2 4 -0.007 1.706
## 3 right_weight 437 3.011 1.407 3 1 5 2 4 -0.015 1.725
## 4 right_distance 437 3.057 1.432 3 1 5 2 4 -0.077 1.668
##
## $data
## left_weight left_distance right_weight right_distance
## 415 4 2 3 5
## 463 4 4 3 3
## 179 2 3 1 4
## 526 5 2 1 1
## 195 2 3 4 5
## 118 1 5 4 3
## 299 3 2 5 4
## 229 2 5 1 4
## 244 2 5 4 4
## 14 1 1 3 4
## 374 3 5 5 4
## 602 5 5 1 2
## 603 5 5 1 3
## 91 1 4 4 1
## 348 3 4 5 3
## 355 3 5 1 5
## 26 1 2 1 1
## 519 5 1 4 4
## 426 4 3 1 1
## 211 2 4 3 1
## 590 5 4 3 5
## 593 5 4 4 3
## 555 5 3 1 5
## 373 3 5 5 3
## 143 2 1 4 3
## 544 5 2 4 4
## 490 4 5 3 5
## 23 1 1 5 3
## 309 3 3 2 4
## 135 2 1 2 5
## 224 2 4 5 4
## 166 2 2 4 1
## 217 2 4 4 2
## 290 3 2 3 5
## 581 5 4 2 1
## 72 1 3 5 2
## 588 5 4 3 3
## 575 5 3 5 5
## 141 2 1 4 1
## 153 2 2 1 3
## 294 3 2 4 4
## 277 3 2 1 2
## 624 5 5 5 4
## 41 1 2 4 1
## 431 4 3 2 1
## 90 1 4 3 5
## 316 3 3 4 1
## 223 2 4 5 3
## 528 5 2 1 3
## 116 1 5 4 1
## 456 4 4 2 1
## 39 1 2 3 4
## 159 2 2 2 4
## 209 2 4 2 4
## 615 5 5 3 5
## 34 1 2 2 4
## 516 5 1 4 1
## 13 1 1 3 3
## 69 1 3 4 4
## 409 4 2 2 4
## 308 3 3 2 3
## 278 3 2 1 3
## 89 1 4 3 4
## 537 5 2 3 2
## 291 3 2 4 1
## 424 4 2 5 4
## 286 3 2 3 1
## 121 1 5 5 1
## 110 1 5 2 5
## 158 2 2 2 3
## 64 1 3 3 4
## 483 4 5 2 3
## 477 4 5 1 2
## 480 4 5 1 5
## 67 1 3 4 2
## 85 1 4 2 5
## 165 2 2 3 5
## 51 1 3 1 1
## 74 1 3 5 4
## 178 2 3 1 3
## 362 3 5 3 2
## 236 2 5 3 1
## 330 3 4 1 5
## 127 2 1 1 2
## 212 2 4 3 2
## 310 3 3 2 5
## 243 2 5 4 3
## 113 1 5 3 3
## 553 5 3 1 3
## 151 2 2 1 1
## 160 2 2 2 5
## 391 4 1 4 1
## 155 2 2 1 5
## 607 5 5 2 2
## 5 1 1 1 5
## 326 3 4 1 1
## 280 3 2 1 5
## 238 2 5 3 3
## 339 3 4 3 4
## 574 5 3 5 4
## 137 2 1 3 2
## 455 4 4 1 5
## 83 1 4 2 3
## 622 5 5 5 2
## 196 2 3 5 1
## 559 5 3 2 4
## 500 4 5 5 5
## 344 3 4 4 4
## 540 5 2 3 5
## 459 4 4 2 4
## 20 1 1 4 5
## 621 5 5 5 1
## 164 2 2 3 4
## 52 1 3 1 2
## 364 3 5 3 4
## 22 1 1 5 2
## 177 2 3 1 2
## 42 1 2 4 2
## 315 3 3 3 5
## 84 1 4 2 4
## 11 1 1 3 1
## 558 5 3 2 3
## 439 4 3 3 4
## 392 4 1 4 2
## 302 3 3 1 2
## 550 5 2 5 5
## 365 3 5 3 5
## 194 2 3 4 4
## 107 1 5 2 2
## 77 1 4 1 2
## 430 4 3 1 5
## 198 2 3 5 3
## 249 2 5 5 4
## 200 2 3 5 5
## 428 4 3 1 3
## 535 5 2 2 5
## 250 2 5 5 5
## 292 3 2 4 2
## 429 4 3 1 4
## 398 4 1 5 3
## 16 1 1 4 1
## 471 4 4 5 1
## 381 4 1 2 1
## 33 1 2 2 3
## 40 1 2 3 5
## 560 5 3 2 5
## 10 1 1 2 5
## 473 4 4 5 3
## 492 4 5 4 2
## 466 4 4 4 1
## 125 1 5 5 5
## 265 3 1 3 5
## 263 3 1 3 3
## 534 5 2 2 4
## 186 2 3 3 1
## 61 1 3 3 1
## 252 3 1 1 2
## 458 4 4 2 3
## 152 2 2 1 2
## 319 3 3 4 4
## 54 1 3 1 4
## 407 4 2 2 2
## 609 5 5 2 4
## 235 2 5 2 5
## 289 3 2 3 4
## 185 2 3 2 5
## 253 3 1 1 3
## 413 4 2 3 3
## 115 1 5 3 5
## 479 4 5 1 4
## 597 5 4 5 2
## 465 4 4 3 5
## 205 2 4 1 5
## 363 3 5 3 3
## 267 3 1 4 2
## 25 1 1 5 5
## 512 5 1 3 2
## 282 3 2 2 2
## 472 4 4 5 2
## 566 5 3 4 1
## 215 2 4 3 5
## 346 3 4 5 1
## 599 5 4 5 4
## 408 4 2 2 3
## 57 1 3 2 2
## 105 1 5 1 5
## 357 3 5 2 2
## 279 3 2 1 4
## 270 3 1 4 5
## 366 3 5 4 1
## 134 2 1 2 4
## 347 3 4 5 2
## 129 2 1 1 4
## 218 2 4 4 3
## 106 1 5 2 1
## 369 3 5 4 4
## 484 4 5 2 4
## 337 3 4 3 2
## 285 3 2 2 5
## 583 5 4 2 3
## 27 1 2 1 2
## 7 1 1 2 2
## 245 2 5 4 5
## 154 2 2 1 4
## 582 5 4 2 2
## 541 5 2 4 1
## 390 4 1 3 5
## 222 2 4 5 2
## 573 5 3 5 3
## 349 3 4 5 4
## 145 2 1 4 5
## 441 4 3 4 1
## 148 2 1 5 3
## 163 2 2 3 3
## 577 5 4 1 2
## 161 2 2 3 1
## 66 1 3 4 1
## 4 1 1 1 4
## 543 5 2 4 3
## 225 2 4 5 5
## 389 4 1 3 4
## 117 1 5 4 2
## 450 4 3 5 5
## 136 2 1 3 1
## 55 1 3 1 5
## 604 5 5 1 4
## 608 5 5 2 3
## 45 1 2 4 5
## 146 2 1 5 1
## 170 2 2 4 5
## 435 4 3 2 5
## 199 2 3 5 4
## 361 3 5 3 1
## 176 2 3 1 1
## 423 4 2 5 3
## 524 5 1 5 4
## 104 1 5 1 4
## 419 4 2 4 4
## 476 4 5 1 1
## 394 4 1 4 4
## 210 2 4 2 5
## 416 4 2 4 1
## 406 4 2 2 1
## 258 3 1 2 3
## 509 5 1 2 4
## 587 5 4 3 2
## 24 1 1 5 4
## 130 2 1 1 5
## 549 5 2 5 4
## 191 2 3 4 1
## 76 1 4 1 1
## 269 3 1 4 4
## 396 4 1 5 1
## 494 4 5 4 4
## 545 5 2 4 5
## 234 2 5 2 4
## 368 3 5 4 3
## 613 5 5 3 3
## 80 1 4 1 5
## 36 1 2 3 1
## 561 5 3 3 1
## 569 5 3 4 4
## 343 3 4 4 3
## 323 3 3 5 3
## 48 1 2 5 3
## 111 1 5 3 1
## 438 4 3 3 3
## 317 3 3 4 2
## 295 3 2 4 5
## 418 4 2 4 3
## 287 3 2 3 2
## 356 3 5 2 1
## 73 1 3 5 3
## 488 4 5 3 3
## 226 2 5 1 1
## 564 5 3 3 4
## 172 2 2 5 2
## 297 3 2 5 2
## 93 1 4 4 3
## 475 4 4 5 5
## 619 5 5 4 4
## 522 5 1 5 2
## 237 2 5 3 2
## 377 4 1 1 2
## 497 4 5 5 2
## 482 4 5 2 2
## 523 5 1 5 3
## 584 5 4 2 4
## 572 5 3 5 2
## 375 3 5 5 5
## 94 1 4 4 4
## 499 4 5 5 4
## 30 1 2 1 5
## 400 4 1 5 5
## 175 2 2 5 5
## 371 3 5 5 1
## 457 4 4 2 2
## 96 1 4 5 1
## 614 5 5 3 4
## 328 3 4 1 3
## 532 5 2 2 2
## 230 2 5 1 5
## 468 4 4 4 3
## 202 2 4 1 2
## 81 1 4 2 1
## 232 2 5 2 2
## 539 5 2 3 4
## 591 5 4 4 1
## 505 5 1 1 5
## 538 5 2 3 3
## 380 4 1 1 5
## 31 1 2 2 1
## 329 3 4 1 4
## 350 3 4 5 5
## 335 3 4 2 5
## 322 3 3 5 2
## 485 4 5 2 5
## 197 2 3 5 2
## 546 5 2 5 1
## 554 5 3 1 4
## 12 1 1 3 2
## 446 4 3 5 1
## 50 1 2 5 5
## 204 2 4 1 4
## 331 3 4 2 1
## 395 4 1 4 5
## 122 1 5 5 2
## 259 3 1 2 4
## 248 2 5 5 3
## 461 4 4 3 1
## 393 4 1 4 3
## 100 1 4 5 5
## 108 1 5 2 3
## 589 5 4 3 4
## 327 3 4 1 2
## 529 5 2 1 4
## 491 4 5 4 1
## 525 5 1 5 5
## 293 3 2 4 3
## 8 1 1 2 3
## 114 1 5 3 4
## 261 3 1 3 1
## 29 1 2 1 4
## 547 5 2 5 2
## 353 3 5 1 3
## 451 4 4 1 1
## 262 3 1 3 2
## 595 5 4 4 5
## 501 5 1 1 1
## 606 5 5 2 1
## 219 2 4 4 4
## 184 2 3 2 4
## 533 5 2 2 3
## 119 1 5 4 4
## 551 5 3 1 1
## 557 5 3 2 2
## 300 3 2 5 5
## 436 4 3 3 1
## 401 4 2 1 1
## 240 2 5 3 5
## 120 1 5 4 5
## 352 3 5 1 2
## 586 5 4 3 1
## 585 5 4 2 5
## 440 4 3 3 5
## 180 2 3 1 5
## 241 2 5 4 1
## 379 4 1 1 4
## 167 2 2 4 2
## 47 1 2 5 2
## 188 2 3 3 3
## 376 4 1 1 1
## 37 1 2 3 2
## 174 2 2 5 4
## 454 4 4 1 4
## 256 3 1 2 1
## 207 2 4 2 2
## 144 2 1 4 4
## 19 1 1 4 4
## 562 5 3 3 2
## 417 4 2 4 2
## 345 3 4 4 5
## 511 5 1 3 1
## 103 1 5 1 3
## 503 5 1 1 3
## 255 3 1 1 5
## 190 2 3 3 5
## 208 2 4 2 3
## 288 3 2 3 3
## 139 2 1 3 4
## 43 1 2 4 3
## 313 3 3 3 3
## 556 5 3 2 1
## 189 2 3 3 4
## 517 5 1 4 2
## 592 5 4 4 2
## 320 3 3 4 5
## 162 2 2 3 2
## 414 4 2 3 4
## 126 2 1 1 1
## 38 1 2 3 3
## 506 5 1 2 1
## 63 1 3 3 3
## 469 4 4 4 4
## 78 1 4 1 3
## 70 1 3 4 5
## 397 4 1 5 2
## 378 4 1 1 3
## 474 4 4 5 4
## 246 2 5 5 1
## 75 1 3 5 5
## 21 1 1 5 1
## 460 4 4 2 5
## 386 4 1 3 1
## 87 1 4 3 2
## 605 5 5 1 5
## 59 1 3 2 4
## 231 2 5 2 1
## 342 3 4 4 2
## 321 3 3 5 1
## 324 3 3 5 4
## 571 5 3 5 1
## 168 2 2 4 3
## 6 1 1 2 1
## 128 2 1 1 3
## 360 3 5 2 5
## 156 2 2 2 1
## 32 1 2 2 2
## 49 1 2 5 4
## 312 3 3 3 2
## 193 2 3 4 3
## 489 4 5 3 4
## 464 4 4 3 4
## 432 4 3 2 2
## 239 2 5 3 4
## 112 1 5 3 2
## 272 3 1 5 2
##
## $subset
## NULL
##
## $outlierMethod
## [1] "none"
##
## attr(,"class")
## [1] "mvn"Menampilkan hasil uji normalitas multivariat.
install.packages("biotools")## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.5'
## (as 'lib' is unspecified)Menginstal package biotools untuk uji homogenitas.
library(biotools)## Loading required package: MASS## ---
## biotools version 4.3Memanggil package biotools.
boxm_result <- boxM(train[, -1], train$class)Melakukan uji homogenitas kovarians menggunakan Box’s M Test.
boxm_result##
## Box's M-test for Homogeneity of Covariance Matrices
##
## data: train[, -1]
## Chi-Sq (approx.) = 104.06, df = 20, p-value = 2.349e-13Menampilkan hasil uji homogenitas kovarians.
install.packages("corrplot")## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.5'
## (as 'lib' is unspecified)Menginstal package corrplot untuk visualisasi korelasi.
library(corrplot)## corrplot 0.95 loadedMemanggil package corrplot.
cor_matrix <- cor(train[, -1])
corrplot(cor_matrix, method = "number")
Menghitung dan menampilkan matriks korelasi antar variabel.
install.packages("car")## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.5'
## (as 'lib' is unspecified)Menginstal package car untuk uji multikolinearitas.
library(car)## Loading required package: carDataMemanggil package car.
model_vif <- lm(left_weight ~ left_distance + right_weight + right_distance, data = train)
vif(model_vif)## left_distance right_weight right_distance
## 1.001310 1.000511 1.001559Menghitung nilai VIF untuk mendeteksi multikolinearitas.
dev.off()## null device
## 1par(mar = c(4,4,2,1))
boxplot(train[, -1], main = "Boxplot Variabel")Menampilkan boxplot untuk mendeteksi outlier pada data.
UJI INDEPENDENSI
contingency_table <- table(train$class, train$left_weight)Membuat tabel kontingensi antara variabel class dan left_weight.
chi_sq_result <- chisq.test(contingency_table)Melakukan uji Chi-Square untuk menguji independensi variabel.
chi_sq_result##
## Pearson's Chi-squared test
##
## data: contingency_table
## X-squared = 84.338, df = 8, p-value = 6.52e-15Menampilkan hasil uji independensi.
ANALISIS DISKRIMINAN(LDA)
install.packages("MASS")## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.5'
## (as 'lib' is unspecified)Menginstal package MASS untuk analisis diskriminan.
library(MASS)Memanggil package MASS.
model_lda <- lda(class ~ ., data = train)Membangun model Linear Discriminant Analysis (LDA).
model_lda## Call:
## lda(class ~ ., data = train)
##
## Prior probabilities of groups:
## B L R
## 0.07551487 0.45308924 0.47139588
##
## Group means:
## left_weight left_distance right_weight right_distance
## B 3.030303 2.969697 3.030303 3.030303
## L 3.580808 3.606061 2.368687 2.454545
## R 2.305825 2.490291 3.626214 3.640777
##
## Coefficients of linear discriminants:
## LD1 LD2
## left_weight 0.6279509 0.50466344
## left_distance 0.5917508 -0.44917839
## right_weight -0.6539825 0.18146589
## right_distance -0.6116543 -0.07225013
##
## Proportion of trace:
## LD1 LD2
## 0.9998 0.0002model_lda$prior## B L R
## 0.07551487 0.45308924 0.47139588model_lda$means## left_weight left_distance right_weight right_distance
## B 3.030303 2.969697 3.030303 3.030303
## L 3.580808 3.606061 2.368687 2.454545
## R 2.305825 2.490291 3.626214 3.640777Menampilkan informasi model LDA, prior, dan rata-rata tiap kelas
pred_lda <- predict(model_lda, test)Melakukan prediksi menggunakan model LDA.
hasil_lda <- pred_lda$classMengambil hasil klasifikasi dari model LDA.
cm_lda <- table(Prediksi = hasil_lda, Aktual = test$class)
cm_lda## Aktual
## Prediksi B L R
## B 0 0 0
## L 12 86 5
## R 4 4 77Membuat confusion matrix untuk evaluasi model LDA.
acc_lda <- mean(hasil_lda == test$class)
acc_lda## [1] 0.8670213Menghitung akurasi model LDA.
REGRESI LOGISTIK MULTINOMIAL
install.packages("nnet")## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.5'
## (as 'lib' is unspecified)Menginstal package nnet untuk regresi logistik multinomial.
library(nnet)Memanggil package nnet.
model_multi <- multinom(class ~ ., data = train)## # weights: 18 (10 variable)
## initial value 480.093570
## iter 10 value 129.927118
## iter 20 value 108.661294
## iter 30 value 108.660701
## final value 108.660693
## convergedMembangun model regresi logistik multinomial.
summary(model_multi)## Call:
## multinom(formula = class ~ ., data = train)
##
## Coefficients:
## (Intercept) left_weight left_distance right_weight right_distance
## L 0.282566 2.487433 2.528075 -2.639700 -2.505189
## R -0.760835 -2.352174 -2.136474 2.392945 2.275570
##
## Std. Errors:
## (Intercept) left_weight left_distance right_weight right_distance
## L 1.045093 0.4731470 0.4711749 0.4852799 0.4656028
## R 1.071423 0.4186127 0.4043911 0.4387926 0.4211602
##
## Residual Deviance: 217.3214
## AIC: 237.3214Menampilkan ringkasan hasil model regresi.
z <- summary(model_multi)$coefficients / summary(model_multi)$standard.errors
p_value <- (1 - pnorm(abs(z), 0, 1)) * 2
p_value## (Intercept) left_weight left_distance right_weight right_distance
## L 0.7868724 1.462577e-07 8.073908e-08 5.341799e-08 7.426758e-08
## R 0.4776320 1.920935e-08 1.269554e-07 4.939459e-08 6.549890e-08Menghitung nilai p-value untuk menguji signifikansi variabel.
pred_multi <- predict(model_multi, test)Melakukan prediksi menggunakan model multinomial.
cm_multi <- table(Prediksi = pred_multi, Aktual = test$class)
cm_multi## Aktual
## Prediksi B L R
## B 4 2 5
## L 9 86 2
## R 3 2 75Membuat confusion matrix untuk model multinomial.
acc_multi <- mean(pred_multi == test$class)
acc_multi## [1] 0.8776596Menghitung akurasi model multinomial.
EVALUASI
install.packages("caret")## Installing package into '/cloud/lib/x86_64-pc-linux-gnu-library/4.5'
## (as 'lib' is unspecified)Menginstal package caret untuk evaluasi model.
library(caret)## Loading required package: ggplot2## Loading required package: latticeMemanggil package caret.
confusionMatrix(hasil_lda, test$class)## Confusion Matrix and Statistics
##
## Reference
## Prediction B L R
## B 0 0 0
## L 12 86 5
## R 4 4 77
##
## Overall Statistics
##
## Accuracy : 0.867
## 95% CI : (0.81, 0.9121)
## No Information Rate : 0.4787
## P-Value [Acc > NIR] : < 2.2e-16
##
## Kappa : 0.754
##
## Mcnemar's Test P-Value : 0.001076
##
## Statistics by Class:
##
## Class: B Class: L Class: R
## Sensitivity 0.00000 0.9556 0.9390
## Specificity 1.00000 0.8265 0.9245
## Pos Pred Value NaN 0.8350 0.9059
## Neg Pred Value 0.91489 0.9529 0.9515
## Prevalence 0.08511 0.4787 0.4362
## Detection Rate 0.00000 0.4574 0.4096
## Detection Prevalence 0.00000 0.5479 0.4521
## Balanced Accuracy 0.50000 0.8910 0.9318Menampilkan evaluasi lengkap model LDA.
confusionMatrix(pred_multi, test$class)## Confusion Matrix and Statistics
##
## Reference
## Prediction B L R
## B 4 2 5
## L 9 86 2
## R 3 2 75
##
## Overall Statistics
##
## Accuracy : 0.8777
## 95% CI : (0.8221, 0.9208)
## No Information Rate : 0.4787
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.7825
##
## Mcnemar's Test P-Value : 0.1752
##
## Statistics by Class:
##
## Class: B Class: L Class: R
## Sensitivity 0.25000 0.9556 0.9146
## Specificity 0.95930 0.8878 0.9528
## Pos Pred Value 0.36364 0.8866 0.9375
## Neg Pred Value 0.93220 0.9560 0.9352
## Prevalence 0.08511 0.4787 0.4362
## Detection Rate 0.02128 0.4574 0.3989
## Detection Prevalence 0.05851 0.5160 0.4255
## Balanced Accuracy 0.60465 0.9217 0.9337Menampilkan evaluasi lengkap model multinomial.
PERBANDINGAN HASIL
cat("Akurasi LDA:", acc_lda, "\n")## Akurasi LDA: 0.8670213cat("Akurasi Multinomial:", acc_multi, "\n")## Akurasi Multinomial: 0.8776596Membandingkan akurasi kedua model.