Table 4 Evaluating the performance of ML models for predicting the facial dimensions from teeth measurements.

From: A novel approach of developing machine learning based models for the prediction of facial dimensions from dental parameters

ML Models

Targets

MSE

RMSE

MAE

Accuracy%

Support Vector Regression (SVR) Model

Al-Al

*0.1150

0.3324

0.2813

91.64%

Ch-Ch

0.6803

0.8248

0.4574

91.64%

Ft-Ft

0.9250

0.9618

0.6941

93.57%

Go-Go

0.7678

0.8762

0.7353

92.81%

Ic -Ic

*0.1190

0.3450

0.2715

90.45%

Oc-Oc

0.5535

0.7440

0.4612

94.31%

Pu-Pu

0.4830

0.6950

0.4189

92.37%

Zy-zy

0.9979

0.9990

0.7034

93.17%

Random Forest

Regression (RFR) Model

Al-Al

*0.1225

0.3499

0.2820

91.65%

Ch-Ch

0.6850

0.8276

0.4615

91.42%

Ft-Ft

0.9040

0.9508

0.6642

93.89%

Go-Go

0.8722

0.9339

0.7453

92.65%

Ic -Ic

*0.1159

0.3404

0.2616

90.72%

Oc-Oc

0.5993

0.7742

0.4832

94.16%

Pu-Pu

0.4618

0.6796

0.4088

92.52%

Zy-zy

0.988

0.9741

0.6472

93.63%

Decision Tree Regression (DTR) Model

Al-Al

0.5186

0.7201

0.4324

86.98%

Ch-Ch

0.7828

0.8848

0.5126

89.87%

Ft-Ft

1.1001

1.0489

0.7555

92.65%

Go-Go

1.3296

1.1531

0.9106

90.05%

Ic -Ic

0.2084

0.4565

0.3507

87.49%

Oc-Oc

0.7363

0.8581

0.5527

92.51%

Pu-Pu

0.5378

0.7333

0.4805

90.54%

Zy-zy

2.1636

1.4709

0.8758

90.23%

Linear Regression (LR) Model

Al-Al

*0.1127

0.3357

0.2781

91.72%

Ch-Ch

0.6709

0.8191

0.4426

91.74%

Ft-Ft

0.9752

0.9875

0.6809

93.73%

Go-Go

0.7978

0.8932

0.6716

93.42%

Ic -Ic

*0.1171

0.3422

0.2596

90.95%

Oc-Oc

0.5533

0.7438

0.4587

94.34%

Pu-Pu

0.4370

0.6611

0.3922

92.81%

Zy-zy

0.9976

0.9988

0.6728

93.54%

  1. bold values without * showing error < 0.5 and bold values with * showing very less error i.e. <0.20.
Back to article page