Page 140 - Proceeding of Atrans Young Researcher's Forum 2019_Neat
P. 140
“Transportation for A Better Life:
Smart Mobility for Now and Then”
23 August 2019, Bangkok, Thailand
Figure 3. Target and predicted output values in a confusion matrix form for the training dataset of:
(a) SVM algorithm; and (b) EDT Bagged algorithm
Figure 4. Target and predicted output algorithm. In order to emphasize the efficiency of
values in a confusion matrix form for the EDT Bagged and SVM, Figure 5 represents the
testing dataset of: (a) SVM algorithm; and errors between target and predicted values for the
(b) EDT Bagged algorithm testing dataset. The value of error criteria such as
RMSE and MAE are also given in the inset
The training dataset is in general dedicated to (RMSE=0.84, MAE=0.34 for SVM; RMSE=0.91,
the construction of AI algorithm, whereas the testing MAE=0.40 for EDT Bagged).
dataset reflects the performance of such given AI
115
