Page 99 - JCAU-5-4

Journal of Chinese
Architecture and Urbanism Cultural landscape in Huizhou City

2.3.4. Cluster analysis of key factors for cultural several different K values were tested for clustering,
landscape as illustrated in Figure 3. When the K value = 5, the
This study employed a combination of qualitative and sum of squares error (SSE) begins to fold, and the
quantitative classification and zoning. Initially, K-means distortion rate slows down. This is considered the
clustering was applied to the seven key factors of the cultural first critical point. Combining this with the natural
landscape in conjunction with descriptive indicators that geographic characteristics of Huizhou City and the
reflect the information. The synthesis of objective and characteristics of the humanities and landscape, it is
accurate cultural landscape zoning of traditional villages in considered that the K=5, representing the number of
Huizhou City was pursued through statistical data analysis clusters for the 132 traditional villages, is the most
and map visualization methods. The calculation steps are reasonable choice.
mainly divided into two parts: (ii) Cluster analysis results of the study villages: Utilizing
(i) Determination of K value (number of clusters): The SPSS 26.0 statistical analysis software, K-means
K-means cluster analysis algorithm required the cluster analysis was conducted on the 7 key factors of
researcher to preset the K value, introducing some the cultural landscape for the 132 study villages with
subjective interference. To avoid interference, the K-value set to 5. This process yielded the values of the
elbow method was adopted in this study to determine cluster centers for the 7 key factors (Table 4). At the
the K value. Using SPSS26.0 on the seven key factors same time, adhering to the principle of the shortest
of the cultural landscape for the 132 research villages, Euclidean distance (3.3), the 132 study sample villages
were classified into 5 distinct classes. These classes
Table 2. Kaiser‑Meyer‑Olkin (KMO) and Bartlett’s were spatially superimposed, laying the foundation for
correlation test the subsequent step in the traditional village cultural
landscape zoning.
KMO value 0.623
Bartlett’s test of sphericity 3. Results
Approximate Chi-square (mathematics) 838.376 3.1. Results of zoning
Degrees of freedom (physics) 231 According to the K-means clustering analysis of
Significance 0.000 traditional villages in Huizhou City, the 132 traditional

Table 3. Matrix of key factor loadings for cultural landscapes
Cultural landscape factor F1 F2 F3 F4 F5 F6 F7
X (River relations) −0.787 −0.206 0.137 0.063 0.010 0.059 0.099
1
X (Village area) 0.718 −0.242 −0.161 0.150 0.203 0.179 −0.174
2
X (Plaza in front of the shrine) 0.647 0.010 0.099 0.215 0.121 −0.399 0.259
3
X (Village site selection) −0.184 −0.844 −0.055 0.000 0.170 −0.075 0.011
4
X (Elevation) −0.280 0.609 0.111 0.144 0.426 0.027 0.135
5
X (Village texture patterns) 0.105 0.179 0.824 −0.155 0.044 0.111 0.015
6
X (Traditional street aspect ratio) −0.193 −0.005 0.723 0.364 0.071 0.067 −0.017
7
X (Traditional street structures) 0.359 0.302 −0.580 −0.226 0.289 −0.085 −0.077
8
X (Water class) 0.200 0.010 0.138 0.789 0.043 0.189 0.010
9
X (Terrain) −0.045 0.605 −0.103 −0.621 0.129 −0.060 −0.106
10
X (Natural landscape to floor area ratio) 0.410 −0.288 0.117 −0.504 −0.033 0.115 0.289
11
X (Village orientation) 0.089 −0.055 0.040 −0.007 0.797 0.065 −0.001
12
X (Feng shui pond) 0.213 0.119 −0.324 0.010 0.523 −0.455 0.021
13
X (Major building forms) −0.106 0.011 0.102 0.173 −0.035 0.849 0.110
14
X (Cuji relationship) 0.383 0.183 0.123 0.041 0.377 0.565 −0.051
15
X (Feng shui forest) 0.148 −0.181 0.116 −0.089 0.244 −0.045 −0.769
16
X (Courtyard) 0.040 −0.137 0.112 −0.094 0.223 0.018 0.725
17
Volume 5 Issue 4 (2023) 7 https://doi.org/10.36922/jcau.1311