Experimental results clearly demonstrate that the generation of an SVM-based classifier system with RFS significantly improves overall classification accuracy as well as producer's and user's accuracies.
Abstract:
The accuracy of supervised land cover classifications depends on factors such as the chosen classification algorithm, adequate training data, the input data characteristics, and the selection of features. Hyperspectral imaging provides more detailed spectral and spatial information on the land cover than other remote sensing resources. Over the past ten years, traditional and formerly widely accepted statistical classification methods have been superseded by more recent machine learning algorithms, e.g., support vector machines (SVMs), or by multiple classifier systems (MCS). This can be explained by limitations of statistical approaches with regard to high-dimensional data, multimodal classes, and often limited availability of training data. In the presented study, MCSs based on SVM and random feature selection (RFS) are applied to explore the potential of a synergetic use of the two concepts. We investigated how the number of selected features and the size of the MCS influence classification accuracy using two hyperspectral data sets, from different environmental settings. In addition, experiments were conducted with a varying number of training samples. Accuracies are compared with regular SVM and random forests. Experimental results clearly demonstrate that the generation of an SVM-based classifier system with RFS significantly improves overall classification accuracy as well as producer's and user's accuracies. In addition, the ensemble strategy results in smoother, i.e., more realistic, classification maps than those from stand-alone SVM. Findings from the experiments were successfully transferred onto an additional hyperspectral data set.
TL;DR: Experimental results based on several hyperspectral image data sets demonstrate that the proposed pixel-pair method can achieve better classification performance than the conventional deep learning-based method.
TL;DR: A comprehensive review of the current-state-of-the-art in DL for HSI classification, analyzing the strengths and weaknesses of the most widely used classifiers in the literature is provided, providing an exhaustive comparison of the discussed techniques.
TL;DR: There are different methods available for classification of imbalance data set which are divided into three main categories, the algorithmic approach, data-preprocessing approach and feature selection approach as mentioned in this paper.
TL;DR: Experimental results with widely used hyperspectral image data sets demonstrate that the proposed classification framework, called diverse region-based CNN, can surpass any other conventional deep learning-based classifiers and other state-of-the-art classifiers.
TL;DR: A new multifeature model, aiming to construct a support vector machine (SVM) ensemble combining multiple spectral and spatial features at both pixel and object levels is proposed, which provides more accurate classification results compared to the voting and probabilistic models.
TL;DR: Internal estimates monitor error, strength, and correlation and these are used to show the response to increasing the number of features used in the forest, and are also applicable to regression.
TL;DR: Setting of the learning problem consistency of learning processes bounds on the rate of convergence ofLearning processes controlling the generalization ability of learning process constructing learning algorithms what is important in learning theory?
TL;DR: This highly anticipated third edition of the most acclaimed work on data mining and machine learning will teach you everything you need to know about preparing inputs, interpreting outputs, evaluating results, and the algorithmic methods at the heart of successful data mining.
TL;DR: Tests on real and simulated data sets using classification and regression trees and subset selection in linear regression show that bagging can give substantial gains in accuracy.
TL;DR: There are several arguments which support the observed high accuracy of SVMs, which are reviewed and numerous examples and proofs of most of the key theorems are given.
Q1. What are the contributions in "Sensitivity of support vector machines to random feature selection in classification of hyperspectral data" ?
The authors investigated how the number of selected features and the size of the MCS influence classification accuracy using two hyperspectral data sets, from different environmental settings.
Q2. What are the common concepts for the construction of classifier ensembles?
Among MCS approaches based on iterative and independent variations of a base classifier, bagging and boosting are perhaps the widest used concepts for the construction of classifier ensembles.
Q3. What is the main reason why the classifier is not stable?
the instability of the base classifier, i.e., a small change in the training samples leads to varying classification results, is an important requirement for a bagging ensemble.
Q4. What is the reason for the large surplus in accuracy achieved by the SVM ensemble approach?
One reason for this might be that, in the case of small suboptimal training sample sets, the SVM classifier is affected by the curse of dimensionality, even though SVMs usually perform well in high-dimensional feature space and with small training sets.
Q5. How many pixels were considered in the experiment?
The scene consists of 145 × 145 pixels, and 14 land cover classes were considered in their experiments, ranging from 54 to 2466 pixels in size.
Q6. What is the way to use the SVM ensemble strategy?
Particularly for small training sample sets, the presented SVM ensemble strategy by RFS constitutes a feasible approach and useful modification of the regular SVM.
Q7. How many features did the SVM ensemble use?
Irrespective of the number of training samples, the use of only 10% of the features (i.e., 16 bands) was ineffective in terms of accuracies (e.g., 93.4% accuracy with 50 iterations and tr#100).
Q8. Why does the use of SVM ensembles yield better results than regular SVM?
Due to the fact that even values outside these ranges yield results superior to those from regular SVM and the relatively small values for ensemble size and feature subset size, the use of SVM ensembles appears worthwhile, and efficient implementation strategies should be investigated.