Ensemble of extreme learning machine for remote sensing image classification

Computer aided diagnosis of schizophrenia on resting state fMRI data by ensembles of ELM

“…• Select disjoint train and test datasets The V-ELM has already been tested in a number of applications, such as hyperspectral image classification (Ayerdi, Marqués, & Graña, 2015), remote sensing data classification (Han & Liu, 2015) and natural gas reservoir characterization (Anifowose, Labadin, & Abdulraheem, in press), wastewater quality index modeling (Zhao, Yuan, Chai, & Tang, 2011), and intrusion detection (Fossaceca, Mazzuchi, & Sarkani, 2015) with the enhancement of multikernel learning. This basic architecture has been modified in the literature, for instance soft-class dependent voting schemes (Cao et al, 2015) provide improved reliability and sparseness of the model, a distributed approach allows to perform classification in P2P networks (Sun, Yuan, & Wang, 2011), and delta test strategy for hidden units selection enhances the construction of ensembles in Yu et al (2014).…”

Induced Subgraph Game for Ensemble Selection

“…The algorithm takes as an input a training set Γ, an ensemble of classifiers Ω, and a selection threshold σ. It begins with computing the ensemble members' predictions P reds of the training samples [lines [3][4][5][6][7], and uses them to build the adjacency matrix w [lines [8][9][10][11][12][13]. Then, it estimates the individual contribution of every classifier using the definition provided by Equation (8) For each h i ∈ Ω…”

A Comparative Study of Data Mining Algorithms for Image Classification

“…Min Han et al [26] found the good classification method which could achieve good classification accuracy to deal with remote sensing image classification. This author proposes a new classification method extreme learning machine (ELM).…”