Meta-XGBoost for Hyperspectral Image Classification Using Extended MSER-Guided Morphological Profiles

Abstract

To investigate the performance of extreme gradient boosting (XGBoost) in remote sensing image classification tasks, XGBoost was first introduced and comparatively investigated for the spectral-spatial classification of hyperspectral imagery using the extended maximally stable extreme-region-guided morphological profiles (EMSER_MPs) proposed in this study. To overcome the potential issues of XGBoost, meta-XGBoost was proposed as an ensemble XGBoost method with classification and regression tree (CART), dropout-introduced multiple additive regression tree (DART), elastic net regression and parallel coordinate descent-based linear regression (linear) and random forest (RaF) boosters. Moreover, to evaluate the performance of the introduced XGBoost approach with different boosters, meta-XGBoost and EMSER_MPs, well-known and widely accepted classifiers, including support vector machine (SVM), bagging, adaptive boosting (AdaBoost), multi class AdaBoost (MultiBoost), extremely randomized decision trees (ExtraTrees), RaF, classification via random forest regression (CVRFR) and ensemble of nested dichotomies with extremely randomized decision tree (END-ERDT) methods, were considered in terms of the classification accuracy and computational efficiency. The experimental results based on two benchmark hyperspectral data sets confirm the superior performance of EMSER_MPs and EMSER_MPs with mean pixel values within region (EMSER_MPsM) compared to that for morphological profiles (MPs), morphological profile with partial reconstruction (MPPR), extended MPs (EMPs), extended MPPR (EMPPR), maximally stable extreme-region-guided morphological profiles (MSER_MPs) and MSER_MPs with mean pixel values within region (MSER_MPsM) features. The proposed meta-XGBoost algorithm is capable of obtaining better results than XGBoost with the CART, DART, linear and RaF boosters, and it could be an alternative to the other considered classifiers in terms of the classification of hyperspectral images using advanced spectral-spatial features, especially from generalized classification accuracy and model training efficiency perspectives.

Combination of Feature Selection and CatBoost for Prediction: The First Application to the Estimation of Aboveground Biomass

“…XGBoost is a boosting algorithm proposed by Chen et al in 2016 based on the GBDT and RF approaches [57]. Compared with GBDT, XGBoost has improvements in multithreaded processing, the classifier, and the optimization function [58]. It also has the following advantages [59,60]:…”

Identification Framework of Contaminant Spill in Rivers Using Machine Learning with Breakthrough Curve Analysis

“…On the other hand, Chen and Guestrin [61] recently suggested XGB to improve the performance of DT. This model is also an ensemble learning method of DT, and appeared as the top model in various machine learning comparison studies [62,63]. The difference from RF is that XGB is based on the gradient boosting method.…”

The Comparison of LightGBM and XGBoost Coupling Factor Analysis and Prediagnosis of Acute Liver Failure

“…From Figure 3, it can be seen that when 'lambda' = 2, the average precision of CV stays at a lower level and remains the same in the interval 'max_ depth'= [2,4]. It starts to increase gradually and reaches its maximum value at 'max_depth' = 6 in [4,6]. When 'lambda' = 3, the average precision of CV stays at a lower level and remains the same in the interval 'max_depth' = [2,6].…”

“…When 'lambda' = 3, the average precision of CV stays at a lower level and remains the same in the interval 'max_depth' = [2,6]. When 'lambda' = 4, the average precision of CV begins to decline gradually and reaches a minimum value at 'max_depth' = 6 in the interval 'max_depth' = [2,4]. It is at a lower level and remains the same level in [4,6].…”

“…Parsa et al applied XGBoost to establish a prediction model for highways, which can detect real-time accident and analyze features [3]. In the remote sensing image field, Samat investigated the performance of XGBoost in remote sensing image classification tasks [4]. In the medical field, Zhang et al compared XGBoost with Artificial Neural Networks (ANN) and Random Forest (RF) in log loss function and training time based on dataset about Acute Liver Failure.…”