psqi_lcga_group_3
HRV_SDNN_MS
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_SDNN_MS']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.692 0.600 0.622 0.633 0.494 0.554 0.486 0.280 0.633 1.000 22.800
LogisticRegression 0.647 0.557 0.599 0.547 0.419 0.441 0.469 0.222 0.547 5.800 17.600
SVM 0.645 0.575 0.607 0.553 0.452 0.456 0.492 0.234 0.553 4.800 15.000
DecisionTree 0.607 0.593 0.597 0.593 0.547 0.589 0.523 0.233 0.593 1.000 16.400
RandomForest 0.594 0.535 0.538 0.547 0.423 0.518 0.406 0.109 0.547 0.400 19.000
XGBoost 0.585 0.553 0.550 0.567 0.430 0.477 0.418 0.149 0.567 0.600 19.400
KNN 0.547 0.466 0.453 0.493 0.308 0.306 0.320 -0.017 0.493 0.600 21.000
MLP 0.483 0.458 0.452 0.467 0.350 0.356 0.347 -0.027 0.467 1.200 18.600
HRV_NN50
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_NN50']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.670 0.592 0.613 0.627 0.481 0.527 0.480 0.270 0.627 1.200 22.800
LogisticRegression 0.637 0.538 0.566 0.533 0.409 0.418 0.459 0.187 0.533 5.200 17.600
SVM 0.627 0.534 0.574 0.513 0.424 0.432 0.494 0.182 0.513 6.000 15.400
XGBoost 0.599 0.563 0.563 0.567 0.485 0.536 0.463 0.168 0.567 0.800 17.600
DecisionTree 0.559 0.536 0.534 0.540 0.440 0.441 0.440 0.128 0.540 1.600 18.000
RandomForest 0.557 0.529 0.532 0.547 0.416 0.514 0.400 0.098 0.547 0.400 20.200
KNN 0.531 0.503 0.490 0.540 0.333 0.334 0.350 0.062 0.540 0.200 22.000
MLP 0.479 0.451 0.443 0.460 0.303 0.298 0.308 -0.040 0.460 0.800 17.800
HRV_PNN50
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_PNN50']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.659 0.586 0.618 0.627 0.464 0.507 0.477 0.279 0.627 1.600 23.200
LogisticRegression 0.624 0.544 0.573 0.540 0.412 0.423 0.463 0.197 0.540 5.200 17.800
SVM 0.611 0.548 0.584 0.527 0.424 0.435 0.468 0.191 0.527 5.400 15.800
XGBoost 0.575 0.561 0.560 0.567 0.482 0.533 0.459 0.163 0.567 0.800 18.200
RandomForest 0.575 0.549 0.576 0.567 0.438 0.693 0.416 0.136 0.567 0.200 20.000
KNN 0.543 0.506 0.493 0.540 0.337 0.337 0.352 0.064 0.540 0.200 21.800
DecisionTree 0.515 0.491 0.490 0.493 0.420 0.433 0.410 0.039 0.493 1.200 17.200
MLP 0.474 0.458 0.450 0.467 0.307 0.304 0.312 -0.027 0.467 1.000 18.200
HRV_RMSSD_MS
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_RMSSD_MS']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.656 0.597 0.623 0.633 0.484 0.535 0.484 0.285 0.633 1.200 23.000
LogisticRegression 0.630 0.551 0.593 0.547 0.415 0.439 0.467 0.215 0.547 5.600 18.200
SVM 0.593 0.549 0.584 0.533 0.408 0.427 0.430 0.185 0.533 5.000 17.200
XGBoost 0.586 0.572 0.571 0.580 0.490 0.543 0.467 0.184 0.580 0.800 18.800
RandomForest 0.562 0.547 0.532 0.567 0.373 0.366 0.382 0.141 0.567 0.200 19.400
KNN 0.549 0.476 0.463 0.513 0.314 0.314 0.330 0.009 0.513 0.400 22.000
DecisionTree 0.542 0.519 0.518 0.520 0.398 0.400 0.397 0.097 0.520 1.400 16.800
MLP 0.471 0.481 0.477 0.487 0.402 0.408 0.399 0.020 0.487 1.400 18.200
HRV_VLF
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_VLF']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.674 0.605 0.637 0.640 0.498 0.567 0.490 0.296 0.640 1.000 23.200
SVM 0.641 0.585 0.612 0.573 0.458 0.465 0.497 0.247 0.573 4.400 17.200
LogisticRegression 0.633 0.568 0.616 0.560 0.427 0.454 0.477 0.245 0.560 5.800 18.000
XGBoost 0.609 0.555 0.555 0.560 0.478 0.529 0.455 0.151 0.560 0.800 18.000
RandomForest 0.600 0.567 0.559 0.587 0.387 0.389 0.394 0.183 0.587 0.600 20.200
KNN 0.588 0.542 0.533 0.567 0.366 0.368 0.375 0.138 0.567 0.600 20.800
DecisionTree 0.502 0.465 0.470 0.460 0.351 0.349 0.354 0.010 0.460 2.200 16.000
MLP 0.487 0.487 0.484 0.493 0.366 0.368 0.367 0.033 0.493 1.600 18.600
HRV_LF
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_LF']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.704 0.615 0.628 0.640 0.507 0.557 0.495 0.295 0.640 1.000 21.800
SVM 0.660 0.600 0.618 0.587 0.457 0.460 0.474 0.261 0.587 3.600 16.600
LogisticRegression 0.654 0.553 0.605 0.540 0.415 0.446 0.465 0.218 0.540 6.200 17.600
RandomForest 0.629 0.603 0.605 0.620 0.469 0.566 0.453 0.250 0.620 0.400 19.800
XGBoost 0.610 0.580 0.580 0.587 0.495 0.546 0.473 0.201 0.587 0.800 18.200
DecisionTree 0.610 0.576 0.580 0.573 0.466 0.463 0.470 0.211 0.573 1.800 16.000
KNN 0.581 0.507 0.492 0.540 0.339 0.336 0.353 0.064 0.540 0.000 21.400
MLP 0.512 0.528 0.524 0.533 0.401 0.407 0.398 0.109 0.533 1.200 17.800
HRV_HF
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_HF']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.676 0.596 0.612 0.627 0.484 0.526 0.482 0.270 0.627 1.200 22.400
LogisticRegression 0.628 0.528 0.567 0.520 0.393 0.412 0.448 0.183 0.520 6.200 17.400
SVM 0.612 0.541 0.571 0.520 0.407 0.417 0.428 0.170 0.520 4.800 15.600
XGBoost 0.593 0.537 0.541 0.547 0.425 0.518 0.408 0.116 0.547 0.400 18.200
RandomForest 0.570 0.537 0.525 0.560 0.363 0.362 0.372 0.123 0.560 0.400 20.400
DecisionTree 0.539 0.509 0.509 0.513 0.411 0.408 0.418 0.081 0.513 2.000 18.400
KNN 0.538 0.463 0.448 0.500 0.303 0.302 0.321 -0.020 0.500 0.400 22.000
MLP 0.425 0.424 0.422 0.427 0.287 0.286 0.288 -0.083 0.427 1.400 17.200
HRV_LF_HF
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_LF_HF']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.702 0.613 0.614 0.633 0.461 0.486 0.457 0.283 0.633 1.000 20.800
SVM 0.668 0.575 0.600 0.567 0.488 0.478 0.568 0.243 0.567 4.600 16.400
LogisticRegression 0.659 0.532 0.564 0.527 0.415 0.421 0.491 0.189 0.527 5.800 17.200
RandomForest 0.656 0.544 0.529 0.567 0.368 0.362 0.378 0.138 0.567 0.200 19.800
XGBoost 0.637 0.566 0.552 0.580 0.387 0.378 0.397 0.182 0.580 0.200 17.800
KNN 0.634 0.550 0.537 0.580 0.372 0.371 0.384 0.151 0.580 0.000 21.200
DecisionTree 0.610 0.590 0.589 0.593 0.497 0.524 0.482 0.226 0.593 1.000 17.200
MLP 0.600 0.577 0.574 0.580 0.478 0.491 0.469 0.202 0.580 1.200 17.400
HRV_POWER
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_POWER']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.700 0.596 0.612 0.627 0.484 0.526 0.482 0.270 0.627 1.200 22.400
SVM 0.644 0.589 0.622 0.567 0.462 0.467 0.502 0.257 0.567 4.800 15.000
RandomForest 0.640 0.550 0.576 0.567 0.438 0.692 0.416 0.138 0.567 0.200 19.800
LogisticRegression 0.632 0.564 0.612 0.547 0.422 0.448 0.471 0.234 0.547 6.200 17.000
XGBoost 0.614 0.584 0.603 0.593 0.517 0.712 0.475 0.200 0.593 0.400 19.200
KNN 0.608 0.532 0.525 0.573 0.354 0.361 0.372 0.130 0.573 0.200 22.600
DecisionTree 0.568 0.547 0.542 0.553 0.419 0.432 0.413 0.142 0.553 1.000 18.000
MLP 0.507 0.470 0.469 0.473 0.318 0.320 0.317 0.005 0.473 1.800 18.000
HRV_EKG_HR
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_EKG_HR']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.663 0.591 0.606 0.627 0.445 0.498 0.445 0.261 0.627 0.800 22.800
LogisticRegression 0.650 0.553 0.596 0.547 0.462 0.466 0.584 0.235 0.547 6.200 16.800
SVM 0.635 0.568 0.607 0.553 0.443 0.460 0.483 0.220 0.553 5.000 17.400
RandomForest 0.610 0.555 0.542 0.573 0.378 0.373 0.386 0.158 0.573 0.400 19.400
KNN 0.590 0.553 0.558 0.573 0.434 0.534 0.418 0.149 0.573 0.400 20.600
XGBoost 0.586 0.550 0.534 0.567 0.377 0.367 0.387 0.148 0.567 0.000 18.200
MLP 0.538 0.484 0.482 0.487 0.405 0.410 0.401 0.028 0.487 1.400 17.400
DecisionTree 0.517 0.499 0.499 0.500 0.427 0.440 0.417 0.056 0.500 1.200 16.800
HRV_EKG_HR_MAXMIN
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_EKG_HR_MAXMIN']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.686 0.597 0.623 0.633 0.484 0.535 0.484 0.285 0.633 1.200 23.000
LogisticRegression 0.648 0.571 0.619 0.560 0.430 0.458 0.478 0.246 0.560 5.800 17.800
SVM 0.611 0.535 0.566 0.520 0.413 0.421 0.458 0.171 0.520 5.200 16.400
RandomForest 0.560 0.524 0.508 0.540 0.356 0.346 0.366 0.095 0.540 0.000 18.200
XGBoost 0.534 0.486 0.479 0.493 0.331 0.325 0.336 0.029 0.493 0.600 17.000
KNN 0.532 0.501 0.486 0.533 0.333 0.331 0.348 0.054 0.533 0.200 21.400
DecisionTree 0.529 0.499 0.499 0.500 0.373 0.372 0.374 0.064 0.500 1.800 17.200
MLP 0.458 0.459 0.454 0.467 0.350 0.356 0.347 -0.023 0.467 1.200 18.200
HRV_RESP_RATE
[資料] 來源=isi_raw_data 目標=psqi_lcga_group_3 列數=150 特徵數=3
[特徵] 使用欄位（前 15）：['BDI_T1', 'BAI_T1', 'HRV_RESP_RATE']
[CV] Stratified 5-fold, seed=42 | class_weight=balanced
[聚合] K-fold = weighted | LOSO = weighted
[Leakage check] Class balance
count percent%
psqi_lcga_group_3
0 84 56.0
1 8 5.3
2 58 38.7
[Leakage check] 未發現與目標 |r| ≥ 0.95 的欄位。
=== Basic ML Benchmark (Stratified 5-fold CV) ===
model AUC F1_pos(=1) Prec_pos Rec_pos F1_neg(=0) Prec_neg Rec_neg MCC Accuracy Pred1_mean Pred0_mean
NaiveBayes 0.680 0.591 0.604 0.620 0.476 0.505 0.478 0.260 0.620 1.400 22.000
LogisticRegression 0.654 0.552 0.617 0.547 0.424 0.463 0.503 0.236 0.547 6.600 18.200
XGBoost 0.620 0.543 0.547 0.553 0.431 0.524 0.413 0.127 0.553 0.400 18.400
SVM 0.615 0.581 0.611 0.567 0.472 0.473 0.533 0.247 0.567 4.800 16.400
RandomForest 0.591 0.562 0.560 0.580 0.435 0.486 0.424 0.169 0.580 0.600 20.200
KNN 0.530 0.470 0.456 0.500 0.313 0.310 0.326 -0.015 0.500 0.200 21.200
MLP 0.500 0.476 0.474 0.480 0.396 0.396 0.397 0.016 0.480 1.600 17.800
DecisionTree 0.495 0.460 0.457 0.467 0.342 0.341 0.346 -0.012 0.467 1.800 18.400