Sparse additive machines (SAMs) have shown competitive performance on variable selection and classification in high-dimensional data due to their representation flexibility and interpretability. However, the existing methods often employ the unbounded or nonsmooth functions as the surrogates of 0-1 classification loss, which may encounter the degraded performance for data with outliers. To alleviate this problem, we propose a robust classification method, named SAM with the correntropy-induced loss (CSAM), by integrating the correntropy-induced loss (C-loss), the data-dependent hypothesis space, and the weighted -norm regularizer ( ) into additive machines. In theory, the generalization error bound is estimated via a novel error decomposition and the concentration estimation techniques, which shows that the convergence rate can be achieved under proper parameter conditions. In addition, the theoretical guarantee on variable selection consistency is analyzed. Experimental evaluations on both synthetic and real-world datasets consistently validate the effectiveness and robustness of the proposed approach.

Generalized additive models (GAMs) have been successfully applied to high dimensional data. However, most existing methods cannot capture the high level feature patterns from complex data. To alleviate this problem, we propose a new sparse additive model, named generalized sparse additive model with unknown link function (GSAMUL), in which the component functions are estimated by B-spline basis and the unknown link function is estimated by a multi-layer perceptron (MLP) network. Furthermore,$\mathscr{l}_{2.1}$-norm regularizer is used for variable selection. The proposed GSAMUL can realize both variable selection and hidden interaction. We integrate this estimation into a bilevel optimization problem, where the data is split into training set and validation set. In theory, we provide the guarantees about the convergence of the approximate procedure. In applications, experimental evaluations on both synthetic and real world data sets consistently validate the effectiveness of GSAMUL.

Due to the high dimensionality of hyperspectral images (HSIs), more training samples are needed in general for better classification performance. However, surface materials cannot always provide sufficient training samples in practice. HSI classification with small size training samples is still a challenging problem. Multiview learning is a feasible way to improve the classification accuracy in the case of small training samples by combining information from different views. This article proposes a new spatial window-based multiview intact feature learning method (SWMIFL) for HSI classification. In the proposed SWMIFL, multiple features that reflect different information of the original image are extracted and spatial windows are imposed on training samples to select unlabeled samples. Then, multiview intact feature learning is performed to learn the intact feature of the training and unlabeled samples. Considering that neighboring samples are likely to belong to the same class, labels of spatial neighboring samples are determined by two factors including the labels of training samples that locate in the spatial window and the labels learned from the intact feature. Finally, unlabeled samples that have same labels under these two factors are treated as new training samples. Experimental results demonstrate that the proposed SWMIFL-based classification method outperforms several well-known HSI classification methods on three real-world data sets.

Many investigations have proved that the acoustics method is intuitive and effective for determining watermelon ripeness. The objective of this work is to drive a new robust acoustics classification scheme KPCA-ELM, which is based on the kernel principal component analysis (KPCA) and extreme learning machine (ELM). Acoustic signals are sampled by a microphone from unripe, ripe and over-ripe watermelon samples, which are randomly divided into two sample sets for training and testing. A set of basic signals is first obtained via KPCA of the training sample. Thus, any given signal can be represented as a linear combination of basis signals, and the coefficients of linear combination are extracted as the features of a signal. Corresponding to the unripe, ripe and over-ripe watermelons, a three-class ELM identification model is constructed based on the training data. The scheme presented in this paper is tested with the testing sample and an accuracy of 92% is achieved. To further evaluate the scheme performance, a comparison of ELM and SVM is conducted in terms of the classification results. The results reveal that the proposed scheme can classify faster than SVM, while ELM is better than SVM in accuracy.

Cross Version Defect Prediction (CVDP) is a practical scenario by training the classification model on the historical data of the prior version and then predicting the defect labels of modules in the current version. Unfortunately, the differences of data distribution across versions may hinder the effectiveness of the trained CVDP model. Thus, it is not trivial to select a suitable training subset from the prior version to promote the CVDP performance. In this paper, we propose a novel method, called Two-Stage Training Subset Selection (TSTSS), to address this challenging issue. In the first stage, TSTSS utilizes a sparse modeling representative selection method to select an initial module subset from the prior version which can well reconstruct the data of the prior version. In the second stage, TSTSS leverages a dissimilarity-based sparse subset selection method to further refine the selected module subset, which enables the selected modules to well represent the modules of the current version. Finally, we use a novel weighted extreme learning machine classifier to construct the CVDP model. We evaluate the CVDP performance of TSTSS on 50 cross-version pairs using 6 indicators. The experiments show that TSTSS can efficiently improve the CVDP performance compared with 11 baseline methods.

Software Defect Data (SDD) are used to build defect prediction models for software quality assurance.Existing work employs feature selection to eliminate irrelevant features in the data to improve prediction performance.Previous studies have shown that different feature selection methods do not always yield similar prediction performance on SDD, which indicates that these methods are not equivalent.Also, previous studies have shown that SDD usually contains noise that may interfere the process of feature selection.In this work, we empirically investigate and measure the equivalence of different feature selection methods for SDD.Further, we intend to analyze the stability of the methods for noisy SDD.We perform statistical analyses on eight projects from NASA dataset with eight feature selection methods.For the equivalence analysis, we introduce Principal Component Analysis (PCA) and overlap index to qualitatively and quantitatively analyze the equivalence of these methods respectively.For the stability analysis, we apply consistency index to measure the stability of these methods.Experimental results indicate that different feature selection methods are indeed not equivalent to each other, and Correlation and Fisher Score methods achieve better stability.