Industrial Efficiency Algorithm Based on Spatio-Temporal-Data-Driven

Table 1

Various weight matrices to be selected.

Expression	Matrix name	Meaning
	Position adjacency weight matrix	The weight matrix is constructed by rook, bishop, and queen position adjacence, and queen adjacence matrix is selected in this paper.; Matrix element
	Geographical distance weight matrix	The weight matrix is constructed by geographical distance between regions, and this paper constructs the weight matrix by reciprocal distance between the centers of provincial capitals in China.; Matrix element , where is the geographical distance between regions and
	Economic distance weight matrix	The weight matrix is constructed by the difference of economic level among different regions, that is, the smaller the economic gap, the stronger the spatial correlation. In this paper, the GDP of each province in China is used as the proxy variable of economic development level to construct the matrix; Matrix element ( represents the annual average GDP of region )
	Adjacency and distance combination weight matrix	Matrix element
	Adjacency and economic combination weight matrix	Matrix element
	Weight matrix of distance and economy combination	Matrix element (where is the geographical distance between region and ; represents the annual average GDP of region )