COMBINED REGIONAL GPS HEIGHT CONVERSION MODEL BASED
ON LEAST SQUARES SUPPORT VECTOR MACHINES
Wang Jigang1，2，3）;Hu Yonghui1） ;and Kong Lingjie4）
1）National Time Service Center，CAS，Xi’an 710600 2）Graduate University of Chinese Academy of Science，Beijing 100039 3）Department of Surveying and Geo-informatics，Huaihai Institute of Technology，Lianyungang 222001 4）College of Geology Engineering and Geomatics，Chang’an University，Xi’an 710054
Abstract：There are many conversion models for GPS height conversion，such as polynomial，BP neural network and multi-quadrics fitting.Because the quasigeoid is an irregular geometric object，every method has both advantages and disadvantages.In order to obtain a more precise and reliable analytical result，the combined model based on least square support vector machines （LSSVM）has been proposed，which has many advantages including fewer parameters and faster training process.As a result，the fitting ability is significantly improved.The combined model still possesses the same important features as the single models.An example is taken and the results are analysed to demonstrate the efficiency of the proposed method.
Wang Jigang,,Hu Yonghui et al. COMBINED REGIONAL GPS HEIGHT CONVERSION MODEL BASED
ON LEAST SQUARES SUPPORT VECTOR MACHINES[J]. jgg, 2009, 29(5): 99-102.
