卷积神经网络和随机森林的城市房价微观尺度制图方法
|
摘要
随着中国城市化进程的加快,城市人口的大规模集聚带来了住房紧张的问题,房价政策制定的时效性与正确性也时刻吸引着社会的关注,因此在微观尺度下对房价进行精细化制图变得愈发重要。由于数据可获取性和现有模型精度的限制,目前已有研究均较少涉及微观尺度。本研究通过将房价数据和遥感影像相融合,构建了一种基于卷积神经网络(CNN)和随机森林(RF)的遥感影像挖掘模型,以实现在不考虑其他数据的情况下,精确、合理地进行房价的微观尺度制图。本文以武汉市作为研究区,在仅有房价数据和遥感影像的情况下,利用本文所构建的模型成功得到武汉市中心城区5 m精度的精细房价图。此外,还利用其他数据源以及挖掘技术与本文所构模型进行了对比分析。结果显示,本文所构建的模型获得了最高的房价模拟拟合优度(R2=0.805),相比传统方法中的最高拟合优度(R2=0.653)其精度提升了23.28%,其制图结果可为政府部门规划决策及武汉市经济分布研究提供基础支撑。
China's rapid urbanization has caused a large number of migrants to move to the city, which has also led to housing shortages. Rapid access to fine-scale house price distribution data plays a very important role in urban housing management, government decision-making, and urban economic model analysis. The availability of data and limitations of existing models make only a few studies involving the mapping of house price distribution at the microscale. By combining house price data with remote sensing images, this study builds a remote sensing image features mining model based on Convolutional Neural Network(CNN) and Random Forest(RF). The proposed CNN-based model in this paper can be applied for accurate and reasonable microscopic mapping of house prices without introducing auxiliary geospatial variables. Only using the house prices data and remote sensing images, we successfully carry out the house prices mapping with the precision of5 meters in the downtown area of Wuhan city. By comparison with the results generated by the other three traditional mining techniques(including A: using spatial datasets extracted from auxiliary geographic dataset only, B: using original features extracted from high-resolution remote sensing images only, C: using original features extracted from high-resolution remote sensing images and auxiliary geographic dataset), the results show that the proposed CNN-based model has the highest house price simulation accuracy(R2=0.805), at least23.28% higher than the fitting accuracies of the traditional methods(A: R2=0.592, B: R2=0.0.434, C: R2=0.653).Moreover, based on the fine-scale house price map, this study further analyzes the spatial heterogeneity distribution of housing prices in the downtown area of Wuhan city. By comparing the partial and overall similarity of the simulated house price distribution map calculated via the perceptual hash algorithm, the results also demonstrate that the house prices distribution of Wuhan city has remarkable fractal characteristics. The micro-scale house price data obtained in this study can provide a basis for microeconomics and fractal research in the urban economics. Meanwhile, this study also provides a brand-new research method for micro-scale economic analysis and resource optimization of large cities in China.
China's rapid urbanization has caused a large number of migrants to move to the city, which has also led to housing shortages. Rapid access to fine-scale house price distribution data plays a very important role in urban housing management, government decision-making, and urban economic model analysis. The availability of data and limitations of existing models make only a few studies involving the mapping of house price distribution at the microscale. By combining house price data with remote sensing images, this study builds a remote sensing image features mining model based on Convolutional Neural Network(CNN) and Random Forest(RF). The proposed CNN-based model in this paper can be applied for accurate and reasonable microscopic mapping of house prices without introducing auxiliary geospatial variables. Only using the house prices data and remote sensing images, we successfully carry out the house prices mapping with the precision of5 meters in the downtown area of Wuhan city. By comparison with the results generated by the other three traditional mining techniques(including A: using spatial datasets extracted from auxiliary geographic dataset only, B: using original features extracted from high-resolution remote sensing images only, C: using original features extracted from high-resolution remote sensing images and auxiliary geographic dataset), the results show that the proposed CNN-based model has the highest house price simulation accuracy(R2=0.805), at least23.28% higher than the fitting accuracies of the traditional methods(A: R2=0.592, B: R2=0.0.434, C: R2=0.653).Moreover, based on the fine-scale house price map, this study further analyzes the spatial heterogeneity distribution of housing prices in the downtown area of Wuhan city. By comparing the partial and overall similarity of the simulated house price distribution map calculated via the perceptual hash algorithm, the results also demonstrate that the house prices distribution of Wuhan city has remarkable fractal characteristics. The micro-scale house price data obtained in this study can provide a basis for microeconomics and fractal research in the urban economics. Meanwhile, this study also provides a brand-new research method for micro-scale economic analysis and resource optimization of large cities in China.
引文
[1]Atalay K,Edwards R,Liu B Y J.Effects of house prices on health:New evidence from Australia[J].Social Science&Medicine,2017,192(21):36-48.
[2]Fan Q.In:Impact of Fluctuation in House Price on Urban Household Consumption[C].Singapore:2ndInternational Conference on Economic,Business Management and Education Innovation,2014:6.
[3]Ge T,Wu T.Urbanization,inequality and property prices:Equilibrium pricing and transaction in the Chinese housing market[J].China Economic Review,2016,45(4):310-328.
[4]Target Range of Chinese House Price Regulation[C].Melbourne:2011 International Conference on Mechanical,Industrial,and Manufacturing Engineering,2011:4.
[5]Liu F Y.Drastic measures for curbing housing price rising[J].China Opening Herald,2006,131(3):11-13,29.
[6]Rondinelli C,Veronese G.Housing rent dynamics in Italy[J].Questioni Di Economia E Finanza,2010,28(62):540-548.
[7]Osland L.An application of spatial econometrics in relation to hedonic house price modeling[J].Journal of Real Estate Research,2010,32(3):289-320.
[8]Duque J C,Patino J E,Ruiz L A,et al.Measuring intra-urban poverty using land cover and texture metrics derived from remote sensing data[J].Landscape&Urban Planning,2015,135(3):11-21.
[9]Huang X,Zhang L.An SVM ensemble approach combining spectral,structural,and semantic features for the classification of high-resolution remotely sensed imagery[J].IEEE Transactions on Geoscience&Remote Sensing,2012,51(1):257-272.
[10]Razavian A S,Azizpour H,Sullivan J,et al.CNN Features off-the-shelf:An astounding baseline for recognition[C].Stockholm:Computer Vision and Pattern Recognition Workshops,2014:512-519.
[11]He J,Xia L,Yao Y,et al.Mining transition rules of cellular automata for simulating urban expansion by using the deep learning techniques[J].International Journal of Geographical Information Science,2018,32(10):1-22.
[12]Yao Y,Zhang J,Hong Y,et al.Mapping fine-scale urban housing prices by fusing remotely sensed imagery and social media data[J].Transactions in GIS,2018,22(2):561-581.
[13]Zhong Y,Fei F,Zhang L.Large patch convolutional neural networks for the scene classification of high spatial resolution imagery[J].Journal of Applied Remote Sensing,2016,10(2):25006.
[14]Tong S,Ling H,Fu Y,et al.Cross-view gait identification with embedded learning[C].New York:Thematic Workshops of Acm Multimedia.ACM,2017.
[15]Krizhevsky A,Sutskever I,Hinton G E.Jeju Island,Korea:ImageNet classification with deep convolutional neural networks[C].Lake Tahoe:Advances in neural information processing systems,2012.
[16]Hinton G E,Srivastava N,Krizhevsky A,et al.Improving neural networks by preventing co-adaptation of feature detectors[J].Computer Science,2012,3(4):212-223.
[17]Simonyan K,Zisserman A.Very deep convolutional networks for large-scale image recognition.arXiv preprint arXiv:1409.1556,2014.
[18]Breiman L.Random forests[J].Machine Learning,2001,45(1):5-32.
[19]詹国旗,杨国东,王凤艳,等.基于特征空间优化的随机森林算法在GF-2影像湿地分类中的研究[J].地球信息科学学报,2018,20(10):1520-1528.[Zhang G Q,Yang G D,Wang F,et al.The random forest classification of wetland from gf-2 imagery based on the optimized feature space[J].Journal of Geo-information Science,2018,20(10):1520-1528.]
[20]Biau G.Analysis of a Random Forests Model[J].Journal of Machine Learning Research,2010,13(2):1063-1095.
[21]单卓然,黄亚平,张衔春.中部典型特大城市人口密度空间分布格局——以武汉为例[J].经济地理,2015,35(9):33-39.[Shan Z R,Huang Y P,Zhang X C.Spatial distribution of population density of typical megacities in central China:A case study of Wuhan[J].Economic Geography,2015,35(9):33-39.]
[22]张俊峰,张安录.土地资源空间异质性与经济发展水平的关系研究——以武汉城市圈为例[J].自然资源学报,2015,30(5):725-735.[Zhang J F,Zhang A L.Study on relationship between spatial heterogeneity of land resources and economic development:A case study of Wuhan urban agglomeration[J].Journal of Natural Resources,2015,30(5):725-735.]
[23]龙莹.空间异质性与区域房地产价格波动的差异——基于地理加权回归的实证研究[J].中央财经大学学报,2010,10(11):80-85.[Long Y.Spatial heterogeneity and the difference of regional fluctuation in housing price:Based on the geographically weighted regression[J].Journal of Central University of Finance&Economics,2010,10(11):80-85.]
[24]何鑫,田丽慧,楚尔鸣.人口流动视角下中国房价波动的空间异质性[J].人口与经济,2017,181(6):43-57.[He X,Tian L H,Chu E M.The spatial heterogeneity of house price volatility in China from the perspective of migration[J].Population&Economics,2017,181(6):43-57.]
[25]Jean N,Burke M,Xie M,et al.Combining satellite imagery and machine learning to predict poverty[J].Science,2016,353(6301):790-794.
[26]Chen X W,Lin X.Big data deep learning:Challenges and perspectives[J].IEEE Access,2014,2:514-525.
[27]Getis A,Ord J K.The analysis of spatial association by use of distance statistics[J].Geographical Analysis,1992,24(3):189-206.
[28]陈涛,刘继生.城市体系分形特征的初步研究[J].人文地理,1994,233(1):25-30.[Chen T,Liu J.Preliminary studies on the fractal properties of systems of towns[J].Human Geography,1994,233(1):25-30.]
[29]Li W,Preneel B.A secure perceptual hash algorithm for image content authentication[J].Lecture Notes in Computer Science,2011,7025(1):108-121.
[30]陈彦光.分形城市与城市规划[J].城市规划,2005,33(2):33-40.[Chen Y G.Fractal cities and city planing[J].City Planning Review,2005,33(2):33-40.]
[31]Wu B,Li R,Huang B.A geographically and temporally weighted autoregressive model with application to housing prices[J].International Journal of Geographical Information Science,2014,28(5):1186-1204.
[32]He C,Wang Z,Guo H,et al.Driving forces analysis for residential housing price in Beijing[J].Procedia Environmental Sciences,2010,2(1):925-936.
[33]Wang L.Empirical study on the impact of urban rail Transit on house rice:Based on hedonic price model[J].Areal Research&Development,2009,28(2):57-61,71.
[34]Wang Y,Wang D L,Liu L H,et al.Spatial differentiation of urban housing prices and its impacts on land market in China[J].China Land Sciences,2015,29(6):33-40.
[35]An H,Yu L,Gupta R.Capital inflows and house prices:Aggregate and regional evidence from China[J].Australian Economic Papers,2016,55(4):451-475.
[36]尹上岗,李在军,宋伟轩,等.基于地理探测器的南京市住宅租金空间分异格局及驱动因素研究[J].地球信息科学学报,2018,20(8):1139.[Yin S G,Li Z J,Song W X,et al.Spatial differentiation and influence factors of residential rent in Nanjing based on geographical detector.Journal of Geo-information Science,2018,20(8):1139-1149.]
[37]郭志懋,周傲英.数据质量和数据清洗研究综述[J].软件学报,2002,13(11):2076-2082.[Guo Z M,Zhou A Y.Research on data quality and data cleaning:A Survey[J].Journal of Software,2002,13(11):2076-2082.]
[38]Chan S.Spatial lock-in:Do falling house prices constrain residential mobility?[J].Journal of Urban Economics,2001,49(3):567-586.
[39]Wen H,Zhang Y,Zhang L.Do educational facilities affect housing price?An empirical study in Hangzhou,China[J].Habitat International,2014,42(42):155-163.
[40]王芳,高晓路,颜秉秋.基于住宅价格的北京城市空间结构研究[J].地理科学进展,2014,33(10):1322-1331.[Wang F,Gao X L,Yan B Q.Research on urban spatial structure in Beijing based on housing prices[J].Progress in Geography,2014,33(10):1322-1331.]
[41]Abadi M,Agarwal A,Barham P,et al.TensorFlow:Largescale machine learning on heterogeneous distributed systems[J].Ar Xiv preprint,ar Xiv:1603.04467,2016-03-14.
[2]Fan Q.In:Impact of Fluctuation in House Price on Urban Household Consumption[C].Singapore:2ndInternational Conference on Economic,Business Management and Education Innovation,2014:6.
[3]Ge T,Wu T.Urbanization,inequality and property prices:Equilibrium pricing and transaction in the Chinese housing market[J].China Economic Review,2016,45(4):310-328.
[4]Target Range of Chinese House Price Regulation[C].Melbourne:2011 International Conference on Mechanical,Industrial,and Manufacturing Engineering,2011:4.
[5]Liu F Y.Drastic measures for curbing housing price rising[J].China Opening Herald,2006,131(3):11-13,29.
[6]Rondinelli C,Veronese G.Housing rent dynamics in Italy[J].Questioni Di Economia E Finanza,2010,28(62):540-548.
[7]Osland L.An application of spatial econometrics in relation to hedonic house price modeling[J].Journal of Real Estate Research,2010,32(3):289-320.
[8]Duque J C,Patino J E,Ruiz L A,et al.Measuring intra-urban poverty using land cover and texture metrics derived from remote sensing data[J].Landscape&Urban Planning,2015,135(3):11-21.
[9]Huang X,Zhang L.An SVM ensemble approach combining spectral,structural,and semantic features for the classification of high-resolution remotely sensed imagery[J].IEEE Transactions on Geoscience&Remote Sensing,2012,51(1):257-272.
[10]Razavian A S,Azizpour H,Sullivan J,et al.CNN Features off-the-shelf:An astounding baseline for recognition[C].Stockholm:Computer Vision and Pattern Recognition Workshops,2014:512-519.
[11]He J,Xia L,Yao Y,et al.Mining transition rules of cellular automata for simulating urban expansion by using the deep learning techniques[J].International Journal of Geographical Information Science,2018,32(10):1-22.
[12]Yao Y,Zhang J,Hong Y,et al.Mapping fine-scale urban housing prices by fusing remotely sensed imagery and social media data[J].Transactions in GIS,2018,22(2):561-581.
[13]Zhong Y,Fei F,Zhang L.Large patch convolutional neural networks for the scene classification of high spatial resolution imagery[J].Journal of Applied Remote Sensing,2016,10(2):25006.
[14]Tong S,Ling H,Fu Y,et al.Cross-view gait identification with embedded learning[C].New York:Thematic Workshops of Acm Multimedia.ACM,2017.
[15]Krizhevsky A,Sutskever I,Hinton G E.Jeju Island,Korea:ImageNet classification with deep convolutional neural networks[C].Lake Tahoe:Advances in neural information processing systems,2012.
[16]Hinton G E,Srivastava N,Krizhevsky A,et al.Improving neural networks by preventing co-adaptation of feature detectors[J].Computer Science,2012,3(4):212-223.
[17]Simonyan K,Zisserman A.Very deep convolutional networks for large-scale image recognition.arXiv preprint arXiv:1409.1556,2014.
[18]Breiman L.Random forests[J].Machine Learning,2001,45(1):5-32.
[19]詹国旗,杨国东,王凤艳,等.基于特征空间优化的随机森林算法在GF-2影像湿地分类中的研究[J].地球信息科学学报,2018,20(10):1520-1528.[Zhang G Q,Yang G D,Wang F,et al.The random forest classification of wetland from gf-2 imagery based on the optimized feature space[J].Journal of Geo-information Science,2018,20(10):1520-1528.]
[20]Biau G.Analysis of a Random Forests Model[J].Journal of Machine Learning Research,2010,13(2):1063-1095.
[21]单卓然,黄亚平,张衔春.中部典型特大城市人口密度空间分布格局——以武汉为例[J].经济地理,2015,35(9):33-39.[Shan Z R,Huang Y P,Zhang X C.Spatial distribution of population density of typical megacities in central China:A case study of Wuhan[J].Economic Geography,2015,35(9):33-39.]
[22]张俊峰,张安录.土地资源空间异质性与经济发展水平的关系研究——以武汉城市圈为例[J].自然资源学报,2015,30(5):725-735.[Zhang J F,Zhang A L.Study on relationship between spatial heterogeneity of land resources and economic development:A case study of Wuhan urban agglomeration[J].Journal of Natural Resources,2015,30(5):725-735.]
[23]龙莹.空间异质性与区域房地产价格波动的差异——基于地理加权回归的实证研究[J].中央财经大学学报,2010,10(11):80-85.[Long Y.Spatial heterogeneity and the difference of regional fluctuation in housing price:Based on the geographically weighted regression[J].Journal of Central University of Finance&Economics,2010,10(11):80-85.]
[24]何鑫,田丽慧,楚尔鸣.人口流动视角下中国房价波动的空间异质性[J].人口与经济,2017,181(6):43-57.[He X,Tian L H,Chu E M.The spatial heterogeneity of house price volatility in China from the perspective of migration[J].Population&Economics,2017,181(6):43-57.]
[25]Jean N,Burke M,Xie M,et al.Combining satellite imagery and machine learning to predict poverty[J].Science,2016,353(6301):790-794.
[26]Chen X W,Lin X.Big data deep learning:Challenges and perspectives[J].IEEE Access,2014,2:514-525.
[27]Getis A,Ord J K.The analysis of spatial association by use of distance statistics[J].Geographical Analysis,1992,24(3):189-206.
[28]陈涛,刘继生.城市体系分形特征的初步研究[J].人文地理,1994,233(1):25-30.[Chen T,Liu J.Preliminary studies on the fractal properties of systems of towns[J].Human Geography,1994,233(1):25-30.]
[29]Li W,Preneel B.A secure perceptual hash algorithm for image content authentication[J].Lecture Notes in Computer Science,2011,7025(1):108-121.
[30]陈彦光.分形城市与城市规划[J].城市规划,2005,33(2):33-40.[Chen Y G.Fractal cities and city planing[J].City Planning Review,2005,33(2):33-40.]
[31]Wu B,Li R,Huang B.A geographically and temporally weighted autoregressive model with application to housing prices[J].International Journal of Geographical Information Science,2014,28(5):1186-1204.
[32]He C,Wang Z,Guo H,et al.Driving forces analysis for residential housing price in Beijing[J].Procedia Environmental Sciences,2010,2(1):925-936.
[33]Wang L.Empirical study on the impact of urban rail Transit on house rice:Based on hedonic price model[J].Areal Research&Development,2009,28(2):57-61,71.
[34]Wang Y,Wang D L,Liu L H,et al.Spatial differentiation of urban housing prices and its impacts on land market in China[J].China Land Sciences,2015,29(6):33-40.
[35]An H,Yu L,Gupta R.Capital inflows and house prices:Aggregate and regional evidence from China[J].Australian Economic Papers,2016,55(4):451-475.
[36]尹上岗,李在军,宋伟轩,等.基于地理探测器的南京市住宅租金空间分异格局及驱动因素研究[J].地球信息科学学报,2018,20(8):1139.[Yin S G,Li Z J,Song W X,et al.Spatial differentiation and influence factors of residential rent in Nanjing based on geographical detector.Journal of Geo-information Science,2018,20(8):1139-1149.]
[37]郭志懋,周傲英.数据质量和数据清洗研究综述[J].软件学报,2002,13(11):2076-2082.[Guo Z M,Zhou A Y.Research on data quality and data cleaning:A Survey[J].Journal of Software,2002,13(11):2076-2082.]
[38]Chan S.Spatial lock-in:Do falling house prices constrain residential mobility?[J].Journal of Urban Economics,2001,49(3):567-586.
[39]Wen H,Zhang Y,Zhang L.Do educational facilities affect housing price?An empirical study in Hangzhou,China[J].Habitat International,2014,42(42):155-163.
[40]王芳,高晓路,颜秉秋.基于住宅价格的北京城市空间结构研究[J].地理科学进展,2014,33(10):1322-1331.[Wang F,Gao X L,Yan B Q.Research on urban spatial structure in Beijing based on housing prices[J].Progress in Geography,2014,33(10):1322-1331.]
[41]Abadi M,Agarwal A,Barham P,et al.TensorFlow:Largescale machine learning on heterogeneous distributed systems[J].Ar Xiv preprint,ar Xiv:1603.04467,2016-03-14.