Research On Policy About Newly-added Construc- Tion Land Quota Controlling Based On Prediction Model And Game Model

Reasonable and scientific control of newly-added construction land is the basic premise to ensure the sustainable utilization of land resources. At present, due to the lack of control measures on newly-added construction land quota process, extensive use of construction land is common on a national scale, which has seriously impacted on the sustainable use of land resources. In order to achieve the goal to protect our limited land resources, this paper proposes a strategy to control the application of newly-added construction land based on a forecasting model and a game theory model. Three key problems are addressed as follows:First, this study introduces modify factors of land use to balance the effects of different industrial factors on construction land scale. Then, we propose a Support Vector Machine-based model and a Restricted Boltzmann Machine-based model to forecast land scales of multiple cities. Considering the characteristics of the construction land data scale, besides utilizing early stopping and weight-delay tricks, we also propose a method based on dropout to train the Restricted Boltzmann Machine-based model for the small scale and continuous data to overcome overfitting during the model training processes.Second, bargaining game model cannot effectively achieve the goal of controlling newly-added construction land quota. This paper proposes an auditing game model for newly-added construction land index control and finally suggests a control strategy based on forecasting results of construction land scale.Finally, local governments’ decision-making behaviors could be affected by the adjacent governments on the dimensionalities of spatial location and government function. So, this paper proposes a new way, which takes both the spatial location and government function into accounts, to analysis this evolutionary game.In the section of case study, this paper uses the construction land data of 8 cities in Zhejiang province from 2008 to 2012 to forecast their construction land scales in 2013. The computational results show that the proposed model can accurately forecast the construction land scales of all cities. Subsequently, this paper simulates the decision-making processes between the superior government and local governments under the strategy considering construction land scale forecasting and local governments’ spatial locations and government functions. Experiments show that, for the superior government, the effects of performing selective-inspection-strategy is better than that of random-inspection-strategy. Finally, we suggest an effective control strategy for controlling newly-added construction land quota based on the forecasting results and game analyses.