Table 1 Analyzing the statistical factors of the developed GPR model with exponential, matern5/2, rational quadratic, and squared exponential kernel functions (Features = 41).

From: Artificial intelligence approach for estimating energy density of liquid metal batteries

Model

Group

R2

MRE

MSE

RMSE

MAE

GPR (Exponential)

Train data

0.997645

0.00839

7.698478

2.774613

1.1020

Test data

0.997492

0.004412

6.302076

2.510394

0.6152

Total data

0.997633

0.007198

7.279962

2.69814

0.9561

GPR (Matern5/2)

Train data

0.995195

0.015379

15.84408

3.980462

2.1249

Test data

0.995793

0.007135

10.35198

3.217449

1.0502

Total data

0.995386

0.012907

14.19724

3.767923

1.8027

GPR (Rational Quadratic)

Train data

0.996497

0.011183

11.52798

3.395288

1.5565

Test data

0.996875

0.005896

7.750085

2.783897

0.7774

Total data

0.996621

0.009598

10.39516

3.224153

1.3229

GPR (Squared Exponential)

Train data

0.992955

0.020191

22.93581

4.789134

2.8428

Test data

0.991203

0.011661

22.65373

4.759594

1.7815

Total data

0.992573

0.017633

22.85122

4.780295

2.5246