Novel Approach to Improve Geocentric Translation Model Performance Using Artificial Neural Network Technology

Ziggah, Yao Yevenyo; Hu, Yincui; Laari, Prosper Basommi; Hui, Zhenyang

doi:10.1590/s1982-21702017000100014

Cited by 8 publications

(8 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From a practical point of view, the testing results produced by the proposed LS-SVM and the widely used BPNN and RBFNN are satisfactory for cadastral surveying applications in Ghana. This statement is buttressed by the ±0.914 m tolerance residual distance set by the Ghana Survey and Mapping Division of Lands Commission for its cadastral surveying works (Yakubu & Kumi-Boateng, 2015;Ziggah, Youjian, Laari, & Hui, 2017). Therefore, the quantitative results presented in Table 4 affirm that assertion.…”

Section: Test Resultsmentioning

confidence: 66%

“…Numerous studies have shown successful implementation of ANN in the geodetic disciplines including but are not limited to deformation studies (Li & Kong, 2014;Huang, Wu, & Ziggah, 2016), meteorological studies (Mohammadi et al, 2015;Durmaz & Karslioglu, 2011), hydrological studies (Tiwari, J. Adamowski, & K. Adamowski, 2016;Deo & Şahin, 2016;Deo, Tiwari, Adamowski, & Quilty, 2017), tidal estimation (Okwuashi & Ndehedehe, 2017), change detection (Pal, 2009;Chang, Han, Yao, Chen, & Xu, 2010), geoid determination (Kavzoglu & Saka, 2005;Sorkhabi, 2015), and gravity field modelling (Turgut, 2016). Additionally, extensive studies on the suitability of ANN for coordinate transformation in both 2D and 3D have also been duly investigated by several authors (Tierra et al, 2008;Zaletnyik, 2004;Lin & Wang, 2006;Tierra, De Freitas, & Guevara, 2009;Tierra & Romero, 2014;Gullu, 2010;Gullu et al, 2011;Turgut, 2010;Mihalache, 2012;Yilmaz & Gullu, 2012;Konakoğlu, Cakir, & Gökalp, 2016;Ziggah, Youjian, Tierra, Konate, & Hui, 2016;Kumi-Boateng & Ziggah, 2017). A thorough review of these coordinate transformation studies indicates that the ANN of radial basis function and backpropagation have been the most commonly used techniques.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Least Squares Support Vector Machine Model for Coordinate Transformation

Ziggah

Issaka

et al. 2019

Geodesy and Cartography

Self Cite

View full text Add to dashboard Cite

In coordinate transformation, the main purpose is to provide a mathematical relationship between coordinates related to different geodetic reference frames. This gives the geospatial professionals the opportunity to link different datums together. Review of previous studies indicates that empirical and soft computing models have been proposed in recent times for coordinate transformation. The main aim of this study is to present the applicability and performance of Least Squares Support Vector Machine (LS-SVM) which is an extension of the Support Vector Machine (SVM) for coordinate transformation. For comparison purpose, the SVM and the widely used Backpropagation Neural Network (BPNN), Radial Basis Function Neural Network (RBFNN), 2D conformal and affine methods were also employed. To assess how well the transformation results fit the observed data, the root mean square of the residual horizontal distances and standard deviation were used. From the results obtained, the LS-SVM and RBFNN had comparable results and were better than the other methods. The overall statistical findings produced by LS-SVM met the accuracy requirement for cadastral surveying applications in Ghana. To this end, the proposed LS-SVM is known to possess promising predictive capabilities and could efficiently be used as a supplementary technique for coordinate transformation.

show abstract

Section: Test Resultsmentioning

confidence: 66%

Section: Introductionmentioning

confidence: 99%

Least Squares Support Vector Machine Model for Coordinate Transformation

Ziggah

Issaka

et al. 2019

Geodesy and Cartography

Self Cite

View full text Add to dashboard Cite

show abstract

“…Notable among them is the three-dimension (3D) conformal transformation models like Bursa-Wolf (Bursa, 1962;Wolf, 1963), Molodensky-Badekas (Molodensky et al, 1962;Badekas, 1969), Abridged Molodensky (Molodensky et al, 1962) and Veis (Veis, 1960). Conversely, in recent times, artificial neural network (ANN) techniques have also been applied for coordinate transformation (Gullu, 2010;Gullu et al, 2011;Lin and Wang, 2006;Mihalache, 2012;Tierra et al, 2008Tierra et al, , 2009Tierra and Romero, 2014;Turgut, 2010;Yilmaz and Gullu, 2012;Zaletnyik, 2004;ElSayed and Ali, 2016;Ziggah et al, 2016;Ziggah et al, 2017a). It was, however, evident in literature that, the statistical test called the hold-out cross-validation has been the most commonly used technique to evaluate coordinate transformation models performance.…”

Section: Introductionmentioning

confidence: 99%

“…Although the data is quite small, the hold-out cross-validation procedure was implemented where 20 co-located points were used to form the model and 7 evenly distributed points were used to test the model. In addition, Ziggah et al (2017a) developed an error compensation model based on ANN capable of improving the performance of the geocentric translation model. Here, 19 co-located points were divided into 11 training points and 8 testing points, respectively.…”

Section: Introductionmentioning

confidence: 99%

Coordinate Transformation between Global and Local Datums Based on Artificial Neural Network with K-Fold Cross-Validation: A Case Study, Ghana

Ziggah

Tierra

et al. 2019

Earth sci. res. j.

Self Cite

View full text Add to dashboard Cite

The popularity of Artificial Neural Network (ANN) methodology has been growing in a wide variety of areas in geodesy and geospatial sciences. Its ability to perform coordinate transformation between different datums has been well documented in literature. In the application of the ANN methods for the coordinate transformation, only the train-test (hold-out cross-validation) approach has usually been used to evaluate their performance. Here, the data set is divided into two disjoint subsets thus, training (model building) and testing (model validation) respectively. However, one major drawback in the hold-out cross-validation procedure is inappropriate data partitioning. Improper split of the data could lead to a high variance and bias in the results generated. Besides, in a sparse dataset situation, the hold-out cross-validation is not suitable. For these reasons, the K-fold cross-validation approach has been recommended. Consequently, this study, for the first time, explored the potential of using K-fold cross-validation method in the performance assessment of radial basis function neural network and Bursa-Wolf model under data-insufficient situation in Ghana geodetic reference network. The statistical analysis of the results revealed that incorrect data partition could lead to a false reportage on the predictive performance of the transformation model. The findings revealed that the RBFNN and Bursa-Wolf model produced a transformation accuracy of 0.229 m and 0.469 m, respectively. It was also realised that a maximum horizontal error of 0.881 m and 2.131 m was given by the RBFNN and Bursa-Wolf. The obtained results per the cadastral surveying and plan production requirement set by the Ghana Survey and Mapping Division are applicable. This study will contribute to the usage of K-fold cross-validation approach in developing countries having the same sparse dataset situation like Ghana as well as in the geodetic sciences where ANN users seldom apply the statistical resampling technique.

show abstract

“…Ziggah et al (2016b) investigated the 3D coordinate transformation performance of the multilayer feed-forward (MLF) neural network, the radial basis function neural network (RBFNN), and multiple linear regression (MLR) and reported that all methods presented satisfactory results. Ziggah et al (2017) proposed a novel approach to improve the geocentric translation model performance based on the MLF neural network, the RBFNN, and the generalized regression neural network (GRNN). The proposed model ANN-ECM (Artificial Neural Network-Error Compensation Model) was found to achieve better transformation accuracy than the geocentric transformation model.…”

Section: Introductionmentioning

confidence: 99%