Page 71 - Çevre Şehir İklim İngilizce - Sayı 4

Hacı Abdullah Uçan - Tayfun Dede - Sinan Nacar

y = 0,6658x + 49764 y = 0,7101x + 46081
300000 R² = 0,6439 300000 R² = 0,574

TreeNet 200000 ANN_15 200000

100000 100000

0 0
0 100000 200000 300000 0 100000 200000 300000
Gerçek daire sayısı
The actual number of flats The actual number of flats
Gerçek daire sayısı
(b)

y = 0,721x + 61669 y = 0,6337x + 73892
300000 300000 R² = 0,5997
R² = 0,7201
TreeNet 200000 ANN_15 200000

100000 100000

0 0
0 100000 200000 300000 0 100000 200000 300000
Gerçek daire sayısı
Gerçek daire sayısı
The actual number of flats The actual number of flats
(c)
Figure 4. Scatter graphs prepared for (a) training, (b) verification and (c)
test data sets of real number of flats and number of flats estimated using
TreeNet and ANN models
When the prepared scatter graphs are examined, it is seen that the data
are closer to the diagonal in the TreeNet method. As can be understood from
these graphs, the model developed using the TreeNet method gave results
closer to the actual number of flats compared to the model developed using
the ANN method.

Results Related to the Number of Licenses

The performance statistics obtained for the training, verification and test
data sets from regression-based KRA, MARS and TreeNet methods established
to estimate the number of licenses variable using mixed selection data sets
are given in Table 7.

60 The Journal of Environment, Urban and Climate