D-1 and only used 
t 5 km away from 
1 we evaluated the 
s. 
of 5 m resolution, 
ate SPOT images. 
y Office, Forestry 
jan. To meet the 
rpolated into 5 x 5 
e system, TWD67 
ansverse Mercator 
1e central meridian 
re acquired from 
esearch, National 
| (O SPOT Image 
System calibration 
> performed on the 
"WD67 Transverse 
m resolution to be 
chose the two-date 
05) because they 
ds less than 10%. 
d from DEM by 
hade data layer by 
and valleys in the 
lo generate terrain 
ys over the study 
ophoto base maps; 
the data layer by 
"he data layer in a 
^w data layer in a 
module, and then 
in position layer 
> derived from the 
1/2005), the other 
ts stated in Hoffer 
(D 
ited using the four 
ing algorithms are 
on 3.3.3 of the free 
and colleagues 
nt/). And other 
sion 1.0 of the free 
inferences from 
), and may remain 
d Stohlgren, 2009). 
| (he concepts of 
d to describe the 
ns, and Bayesian 
istribution of each 
  
pixel when the entropy reach the maximum that the state would 
be extremely close to uniform distribution. That is, 
MAXENT would find out the type of probability distribution 
that is most likely occurring in the general state. The formula 
for MAXENT is shown in following equation (2): 
  
P. f (x)-min : : ; 
= A *- — linearPredictorNormalizer | /Z 
Kum 2 " max, — min, If (2) 
x)-mi : 
where Aoi mig 1 = hinge feature 
max-min 
À, = weight coefficient 
linear predictor normalize = a constant for numerical 
stability 
Z: a scaling constant that ensures that P sums to 1 
over all grid cells 
2) DOMAIN derives a point-to-point similarity metric to assign 
a classification value to a potential site based on its proximity 
in environmental space to the most similar occurrence. The 
Gower metric (Gower, 1971) provides a suitable means of 
quantifying similarity between two sites. The distance of d 
between two points A and B in a Euclidean p dimensional space 
is defined as equation (3): 
14A, -B,| 
d E. k k 3 
4 a Rangek ) G) 
We define the complementary similarity measure R 45: 
Rjp71- dig (4) 
R is constrained between 0 and 1 for points within the ranges 
use in Equation 3, 
We define S4, the maximum similarity between candidate point 
Aand the set of known record sites Tj as equation (5): 
Sy =max Ry, (5) 
Jel 
By evaluating S for all grid points in a target area, a matrix of 
continuous varying similarity values is generated which are not 
probability estimates, but degrees of classification confidence 
(Carpenter er al., 1993). 
3) GLM is a generalization of general linear models. General 
Class of linear models are made up of three components: 
random, systematic, and link function. Random component 
identifies response variable E(Y) and its probability distribution. 
Systematic component identifies the set of predictor variables 
(M. Link function identifies a function of the mean that 
Is a linear function g(u) of the predictor variables. The 
formula for GLM is shown in following equation (3): 
Eos sua p X, v BL, (6) 
      
      
    
  
  
  
   
      
    
    
     
     
   
    
  
    
    
    
     
    
   
   
  
   
     
  
   
    
  
   
       
  
    
     
  
    
   
    
  
  
  
  
where d - constants 
B - regression coefficients 
X — predictor variable 
By using a logit link function that transforms the scale of the 
response variable, being able to relax the distribution and 
constancy of variances assumptions that are commonly 
required by traditional linear models (McCullagh and Nedler, 
1989). Consequently, the GLM model is particularly suitable 
for predicting species distributions, and has been proven to be 
successful in various ecological applications (Guisan ef al., 
2002). 
4) Back-propagation artificial neural network (BPANN) 
consists of input, hidden, and output layers. The input layer 
may contain information about individual training pixels 
including percent spectral reflectance in various bands and 
ancillary data such as elevation, slope, etc. 
Each layer consists of nodes that are interconnected. This 
interconnectedness allows information to flow in multiple 
directions as the network is trained. The weight of these 
interconnections is eventually learned by the neural network 
and stored. These weights are used during the output layer 
might represent a single thematic map land-cover class. 
We set four layers (one input layer, one output layer, and two 
hidden layers) that can be trained using back propagation 
algorithm and particle swarm optimization (PSO) algorithm is 
implement. The structure of back propagation neural network 
is shown in figure 2. 
  
input values 
    
output values 
  
  
  
Figure 2. The structure of back propagation 
artificial neural network 
3.4 Model Validation 
Evaluation methods of the different samplings, we used 
split-sample validation. The first one (training dataset) be 
used to build model; the other one (test dataset) be used to 
validate the model. For each model, predicted the response of 
the remaining data, and calculated the error matrix (De'ath and 
Fabricius, 2000). Some common statistical measurements 
included producer's accuracy, user's accuracy, overall accuracy 
and Kappa coefficient (Jensen, 2005; Lillesand et al., 2008). 
4. RESULSTS AND DISCUSSION 
Initially, we depicted and compared the effect of micro-terrain 
feature in two watersheds as shown in table 1 and figure 3. 
The Tong-Feng watershed has not only steep valley but also