Prediction of geoid undulations: Random forest versus classic interpolation techniques

Abstract

Local geoid determination studies are commonly carried out today to establish the relationship between the ellipsoidal height (h)$$ (h) $$ determined by satellite geodesy methods and the orthometric height (H)$$ (H) $$ found using geoid undulation (N)$$ (N) $$. The aim of this study was to determine a local geoid using the kriging, local polynomial (LP), and inverse distance to a power (IDP) interpolation methods along with the random forest (RF) regression method and to compare the performance. For the application, 193 Global Navigation Satellite System (GNSS)/leveling points homogeneously distributed over the study area were selected as reference data, and 70 GNSS/leveling points were selected as test data. The results were compared in terms of accuracy using well‐known performance metrics, namely, the root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2) values. According to the results, the highest R2 and the lowest RMSE and MAE values were obtained with the RF regression method. These findings demonstrated the superiority of the RF regression over the classic interpolation methods applied in the local geoid determination for the study area.

Random forest method for estimation of brake specific fuel consumption

“…Compared to other machine learning methods, this method has low computation requirements and high precision and is not sensitive to multicollinearity. It also demonstrates good robustness to missing and unbalanced data 10 – 13 . In this regard, an improved RF method has been introduced in this study to enhance the accuracy of BSFC estimation, marking the first application of this method in estimating the BSFC map.…”

“…Common calculation methods include the K-nearest neighbor (KNN) method, polynomial regression, inverse distance weighted (IDW) method, ordinary kriging (OK) method, and multi-layer perceptron (MLP) method. However, these methods are known to have large errors in estimating uniformly distributed data 10 . This is particularly problematic when drawing high-resolution prediction maps for weather data 11 , where accuracy is essential.…”

Geodetic Data Processing Using Soft Computing Techniques

“…They are particularly effective in pattern recognition, learning from data, and optimizing complex problems. The flexibility and adaptability of soft computing techniques have made them an attractive alternative to traditional geodetic data processing methods, often leading to improved performance and more accurate results [10,11]. In addition, soft computing techniques are often employed in conjunction with traditional geodetic data processing methods, resulting in hybrid approaches that leverage the strengths of both paradigms [12].…”