The Evaluation of Accuracy Performance in an Enhanced Embedded Feature Selection for Unstructured Text Classification

Feature selection and analysis on correlated gas sensor data with recursive feature elimination

Prediction of the stage of cancer plays an important role in planning the course of treatment and has been largely reliant on imaging tools which do not capture molecular events that cause cancer progression. Gene-expression data-based analyses are able to identify these events, allowing RNA-sequence and microarray cancer data to be used for cancer analyses. Breast cancer is the most common cancer worldwide, and is classified into four stages - stages 1, 2, 3, and 4 [2]. While machine learning models have previously been explored to perform stage classification with limited success, multi-class stage classification has not had significant progress. There is a need for improved multi-class classification models, such as by investigating deep learning models. Gene-expression-based cancer data is characterised by the small size of available datasets, class imbalance, and high dimensionality. Class balancing methods must be applied to the dataset. Since all the genes are not necessary for stage prediction, retaining only the necessary genes can improve classification accuracy. The breast cancer samples are to be classified into 4 classes of stages 1 to 4. Invasive ductal carcinoma breast cancer samples are obtained from The Cancer Genome Atlas (TCGA) and Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) datasets and combined. Two class balancing techniques are explored, synthetic minority oversampling technique (SMOTE) and SMOTE followed by random undersampling. A hybrid feature selection pipeline is proposed, with three pipelines explored involving combinations of filter and embedded feature selection methods: Pipeline 1 - minimum-redundancy maximum-relevancy (mRMR) and correlation feature selection (CFS), Pipeline 2 - mRMR, mutual information (MI) and CFS, and Pipeline 3 - mRMR and support vector machine-recursive feature elimination (SVM-RFE). The classification is done using deep learning models, namely deep neural network, convolutional neural network, recurrent neural network, a modified deep neural network, and an AutoKeras generated model. Classification performance post class-balancing and various feature selection techniques show marked improvement over classification prior to feature selection. The best multiclass classification was found to be by a deep neural network post SMOTE and random undersampling, and feature selection using mRMR and recursive feature elimination, with a Cohen-Kappa score of 0.303 and a classification accuracy of 53.1%. For binary classification into early and late-stage cancer, the best performance is obtained by a modified deep neural network (DNN) post SMOTE and random undersampling, and feature selection using mRMR and recursive feature elimination, with an accuracy of 81.0% and a Cohen-Kappa score (CKS) of 0.280. This pipeline also showed improved multiclass classification performance on neuroblastoma cancer data, with a best area under the receiver operating characteristic (auROC) curve score of 0.872, as compared to 0.71 obtained in previous work, an improvement of 22.81%. The results and analysis reveal that feature selection techniques play a vital role in gene-expression data-based classification, and the proposed hybrid feature selection pipeline improves classification performance. Multi-class classification is possible using deep learning models, though further improvement particularly in late-stage classification is necessary and should be explored further.

Cancer is a world-famous deadly disease. According to the World Health Organization (WHO), cancer is the second leading cause of death globally and is responsible for an estimated 9.6 million deaths in 2018. One well-known technique for cancer detection is the DNA microarray technique. DNA microarray technology provides an opportunity for researchers to analyze thousands of gene expression profiles at the same time to determine whether a person has cancer or not. However, one of the problems in DNA microarray data is the large number of features that require feature selection. In overcoming these problems, this study will use the feature selection Support Vector Machine-Recursive Feature Elimination (SVM-RFE) and Chi-Square and use the Naive Bayes classification method. The accuracy results from using feature selection with those that are not will be compared. The accuracy between using the two feature selection methods will also be compared to find which feature selection method is better when combined with the Naive Bayes classification method. To get an overall picture of the performance comparison, this study also considers precision, recall, and F1-score. The best accuracy results obtained were 100% lung cancer data with SVM-RFE and Chi-Square, 99.6% ovarian cancer with SVM-RFE, 93.7% breast cancer with SVM-RFE, and 90% colon cancer with SVM- RFE.

Effects of imaging modalities, brain atlases and feature selection on prediction of Alzheimer's disease

As a feature selection method, support vector machines-recursive feature elimination (SVM-RFE) can remove irrelevance features but don’t take redundant features into consideration. In this paper, it is shown why this method can’t remove redundant features and an improved technique is presented. Correlation coefficient is introduced to measure the redundancy in the selected subset with SVM-RFE. The features which have a great correlation coefficient with some important feature are removed. Experimental results show that there actually are several strongly redundant features in the selected subsets by SVM-RFE. The coefficients are high to 0.99. The proposed method can not only reduce the number of features, but also keep the classification accuracy.

• Neural Network feature selection methodology for both numerical and categorical data. • Embedded feature selection provides model and feature subset in a single training. • Novel feature embedding for dealing with high-cardinality categorical features. • Validation of the methodology on several open and real industry dataset. • Comparison with several state-of-the-art feature selection methodologie. In an era of effortless data collection, the impact of machine learning — especially neural networks (NNs) — is undeniable. As datasets grow in size and complexity, efficiently handling mixed data types, including categorical and numerical features, becomes critical. Feature encoding and selection play a key role in improving NN performance, efficiency, interpretability, and generalisation. This paper presents GLEm-Net (Grouped Lasso with Embeddings Network), a novel NN-based approach that seamlessly integrates feature encoding and selection directly into the training process. GLEm-Net uses embedding layers to process categorical features with high cardinality, simplifying the model and improving generalisation. By extending the grouped Lasso regularisation to explicitly consider categorical features, GLEm-Net automatically identifies the most relevant features during training and returns them to the analyst. We evaluate GLEm-Net on open and proprietary industry datasets and compare it to state-of-the-art feature selection methodologies. Results show that GLEm-Net adapts to each dataset by allowing the NN to directly select subsets of most important features, offering on par performance with the best state-of-the-art feature selection methods, while eliminating the need for the external feature encoding and selection steps that are now incorporated in the NN training stage.

Improved multiclass feature selection via list combination

In a DNA microarray dataset, gene expression data often has a huge number of features(which are referred to as genes) versus a small size of samples. With the development of DNA microarray technology, the number of dimensions increases even faster than before, which could lead to the problem of the curse of dimensionality. To get good classification performance, it is necessary to preprocess the gene expression data. Support vector machine recursive feature elimination (SVM-RFE) is a classical method for gene selection. However, SVM-RFE suffers from high computational complexity. To remedy it, this paper enhances SVM-RFE for gene selection by incorporating feature clustering, called feature clustering SVM-RFE (FCSVM-RFE). The proposed method first performs gene selection roughly and then ranks the selected genes. First, a clustering algorithm is used to cluster genes into gene groups, in each which genes have similar expression profile. Then, a representative gene is found to represent a gene group. By doing so, we can obtain a representative gene set. Then, SVM-RFE is applied to rank these representative genes. FCSVM-RFE can reduce the computational complexity and the redundancy among genes. Experiments on seven public gene expression datasets show that FCSVM-RFE can achieve a better classification performance and lower computational complexity when compared with the state-the-art-of methods, such as SVM-RFE.

Suicidal attempts in patients with major depressive disorder (MDD) have become an important challenge in global mental health affairs. To correctly distinguish MDD patients with and without suicidal attempts, a multimodal prediction model was developed in this study using multimodality data, including demographic, depressive symptoms, and brain structural imaging data. This model will be very helpful in the early intervention of MDD patients with suicidal attempts. Two feature selection methods, support vector machine-recursive feature elimination (SVM-RFE) and random forest (RF) algorithms, were merged for feature selection in 208 MDD patients. SVM was then used as a classification model to distinguish MDD patients with suicidal attempts or not. The multimodal predictive model was found to correctly distinguish MDD patients with and without suicidal attempts using integrated features derived from SVM-RFE and RF, with a balanced accuracy of 77.78%, sensitivity of 83.33%, specificity of 70.37%, positive predictive value of 78.95%, and negative predictive value of 76.00%. The strategy of merging the features from two selection methods outperformed traditional methods in the prediction of suicidal attempts in MDD patients, with hippocampal volume, cerebellar vermis volume, and supracalcarine volume being the top three features in the prediction model. This study not only developed a new multimodal prediction model but also found three important brain structural phenotypes for the prediction of suicidal attempters in MDD patients. This prediction model is a powerful tool for early intervention in MDD patients, which offers neuroimaging biomarker targets for treatment in MDD patients with suicidal attempts.

Skin cancer is on the rise. Hence the accurate detection of cancerous lesions is of paramount importance in the treatment of this health condition. In this study, we present a computer vision framework that studies 10 skin lesion features including newly introduced morphological and texture features along with established in the literature. All features are extracted automatically from 90 lesion images. A two-class classification problem was applied to determine the most significant features of disease using three features selection methods, Support Vector Machines Recursive Feature Elimination (SVMRFE), Information Gain, and Correlation-based Feature Subset Selection. We found that the five features selected by SVMRFE provides the highest accuracy readings of 100% model, 84% leave-one-out, and 89% 10-fold cross validation (10×CV) than the other two methods. Comparing the selected features with those used by the Total Dermoscopy Score, we report consistencies, disagreements, and the contributions of the present work.

Recursive feature elimination with random forest for PTR-MS analysis of agroindustrial products

Deep belief network-based approach for detecting Alzheimer's disease using the multi-omics data

IntroductionIn the classification of Mass Spectrometry (MS) proteomics data, peak detection, feature selection, and learning classifiers are critical to classification accuracy. To better understand which methods are more accurate when classifying data, some publicly available peak detection algorithms for Matrix assisted Laser Desorption Ionization Mass Spectrometry (MALDI-MS) data were recently compared; however, the issue of different feature selection methods and different classification models as they relate to classification performance has not been addressed. With the application of intelligent computing, much progress has been made in the development of feature selection methods and learning classifiers for the analysis of high-throughput biological data. The main objective of this paper is to compare the methods of feature selection and different learning classifiers when applied to MALDI-MS data and to provide a subsequent reference for the analysis of MS proteomics data.ResultsWe compared a well-known method of feature selection, Support Vector Machine Recursive Feature Elimination (SVMRFE), and a recently developed method, Gradient based Leave-one-out Gene Selection (GLGS) that effectively performs microarray data analysis. We also compared several learning classifiers including K-Nearest Neighbor Classifier (KNNC), Naïve Bayes Classifier (NBC), Nearest Mean Scaled Classifier (NMSC), uncorrelated normal based quadratic Bayes Classifier recorded as UDC, Support Vector Machines, and a distance metric learning for Large Margin Nearest Neighbor classifier (LMNN) based on Mahanalobis distance. To compare, we conducted a comprehensive experimental study using three types of MALDI-MS data.ConclusionRegarding feature selection, SVMRFE outperformed GLGS in classification. As for the learning classifiers, when classification models derived from the best training were compared, SVMs performed the best with respect to the expected testing accuracy. However, the distance metric learning LMNN outperformed SVMs and other classifiers on evaluating the best testing. In such cases, the optimum classification model based on LMNN is worth investigating for future study.

Hyperspectral images provide important information for addressing complex classification problems required for a detailed characterization of spectral behavior of the target objects. Classification of such datasets into meaningful land use and land cover classes (LULC) has been the most concentrated topic in remote sensing arena. Rotation forest (RotFor), a new ensemble learning method, has been recently proposed as an alternative to conventional classifiers in the context of multispectral and hyperspectral image classification. In this study, the use of RotFor was investigated for the classification of hyperspectral imagery, specifically an AVIRIS image data. Support vector machine (SVM) was also used as a benchmark classifier. In order to select the best contributing bands of AVIRIS, support vector machine-recursive feature elimination (SVM-RFE) approach was applied. Results of this study showed that RotFor algorithm provided more accurate classification results than the SVM classifier with the use of smaller size data sets selected by SVM-RFE. Based on the Wilcoxon's signed-rank test, the performance difference between RotFor and SVM was found to be statistically significant.

During image classification, it is important to select optimal bands in order to reduce the amount of information while maintaining a considerable accuracy. However, problems exist in the artificial subjective selection of the bands, and as a result, the true accuracies of the classifications are not effectively expressed. This research was conducted using 10 bands of the Landsat-8 fusion data as the information sources. The maximum likelihood classification–recursive feature elimination (MLC-RFE) method and the support vector machine-recursive feature elimination (SVM-RFE) method were used to select the most appropriate bands, and to analyse the classification results of different methods. The results showed that the bands obtained from MLC and SVM were different, with MLC being more sensitive to the change in features compared with SVM. The highest and lowest overall accuracies of MLC were 99.1667% and 92.2312%, respectively (the corresponding Kappa coefficients were 0.9903 and 0.9094). The highest and lowest overall accuracies of SVM were 98.7231% and 86.6398%, respectively (the corresponding Kappa coefficients were 0.9851 and 0.8441). The two classifiers differed greatly as regards the classification accuracies of the optimal and worst band combinations. This indicates that the selection of the participating bands was very important for the image classification, and MLC-RFE is more suitable than SVM-RFE for Landsat-8 image classification.