Table 5 Rank of ML Models based on Prediction Error Assessment Using Datasets Obtained between 4 and 5 December 2021.

From: Comparative study of ten machine learning algorithms for short-term forecasting in gas warning systems

Model

MAE

Rank

MSE

Rank

RMSE

Rank

SAE

Rank

Average of Rank

Rank_Prediction Errors

RF

0.004849

1

0.000087

1

0.009320

1

22.304

1

1

1

ARIMA

0.005011

2

0.000087

1

0.009341

2

23.05

2

1.75

2

LR

0.005011

2

0.000087

1

0.009341

2

23.05

2

1.75

2

SVM

0.006651

4

0.000094

4

0.009719

4

30.594

4

4

4

KNN

0.009780

5

0.000166

5

0.012870

5

44.99

5

5

5

RNN

0.056621

7

0.005646

6

0.075140

6

559.08

7

6.5

6

LSTM

0.055999

6

0.016717

9

0.129290

9

545.58

6

7.5

7

BP_Resilient

0.056660

8

0.005659

7

0.075229

7

1,198,900

9

7.75

8

BP_SOG

0.056770

9

0.005685

8

0.075398

8

1,201,300

10

8.75

9

Perceptron

0.878530

10

0.774710

10

0.880180

10

4600

8

9.5

10

Back to article page