Regression-based Feature Selection Research Articles

An amalgamated correlation and regression based feature selection with ensemble learning approach for IoT network attack detection

AbstractThe advancements in the field of Information and Communication Technology (ICT) have led to the development of the Internet of Things (IoT), where ordinary things (e.g. smart meters, sensors etc.) can be connected and controlled over the internet. However, an IoT network is vulnerable to many malicious network attacks due to its inherent properties like limited computational abilities, heterogeneity, massive connectivity, etc. This paper presents a feature selection technique, which uses an amalgamation of LASSO regression and correlation based techniques, to find appropriate features for IoT network attack detection. The effectiveness of the proposed mechanism was evaluated on benchmark IoT network datasets using different machine learning techniques. The results revealed that the Gradient Boosting ensemble learning model achieved a maximum attack detection accuracy of 99.98%, and outperformed the other studied models.

Sparse orthogonal supervised feature selection with global redundancy minimization, label scaling, and robustness

Selecting discriminative features to build effective learning models is a significant research work in machine learning. In practical applications, the data distribution characteristics are diverse, and the uncertainties pose challenges for building learning models with robustness and generalization capabilities. Since one-hot encoding is good at representing independent labels, the label matrix of regression-based feature selection (FS) methods is usually encoded with one-hot encoding. However, it's not well adapted to the different data distributions. This paper proposes a sparse orthogonal supervised FS model with global redundancy minimization, label scaling, and robustness (GRMLSRSOFS) to address the above problems. This model uses the label scaling technique proposed in this paper to better adapt to different data distributions. An iterative optimization method is given, and its convergence is demonstrated theoretically and experimentally. Further, experimental results on 12 public datasets show that 1) The GRMLSRSOFS can achieve higher classification accuracy with fewer features in most cases than several state-of-the-art FS methods. For example, the GRMLSRSOFS achieves 100% classification accuracy using only 20 features on the warpPIE10P dataset and obtains nearly 6% improvement over other methods on the Yale dataset. 2) The convergence speed of the GRMLSRSOFS will be faster after label scaling.

Performance Analysis and Assessment of Type 2 Diabetes Screening Scores in Patients with Non-Alcoholic Fatty Liver Disease

Type 2 diabetes (T2D) and non-alcoholic fatty liver disease (NAFLD) are worldwide chronic diseases that have strong relationships with one another and commonly exist together. Type 2 diabetes is considered one of the risk factors for NAFLD, so its occurrence in people with NAFLD is highly likely. As the high and increasing number of T2D and NAFLD, which potentially followed by existing together number, an analysis and assessment of T2D screening scores in people with NAFLD is necessary to be done. To prevent this potential case, an effective early prediction model is also required to be developed, which could help the patients avoid the dangers of both existing diseases. Therefore, in this study, analysis and assessment of T2D screening scores in people with NAFLD and the early prediction model utilizing a forward logistic regression-based feature selection method and multi-layer perceptrons are proposed. Our analysis and assessment results showed that the prevalence of T2D among patients with NAFLD was 8.13% (for prediabetes) and 37.19% (for diabetes) in two population-based NAFLD datasets. The variables related to clinical tests, such as alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), gamma-glutamyl transferase (GGT), and systolic blood pressure (SBP), were found to be statistically significant predictors (p-values < 0.001) that indicate a strong association with T2D among patients with NAFLD in both the prediabetes and diabetes NAFLD datasets. Finally, our proposed model showed the best performance in terms of all performance evaluation metrics compared to existing various machine learning models and also the models using variables recommended by WHO/CDC/ADA, with achieved accuracy as much as 92.11% and 83.05% and its improvement scores after feature selection of 1.35% and 5.35%, for the first and second dataset, respectively.

Design and Implementation of ML Model for Early Diagnosis of Parkinson’s Disease using Gait Data Analysis in IoT Environment

Parkinson’s disease (PD) is the world’s second most neurodegenerative disorder that results in a steady loss of movement. The symptoms in patients occur slowly with the passage of time are and very hard to identify in its initial stage. So, early diagnosis of PD is the foremost need for timely treatment to people. The introduction of smart technologies like the Internet of Things (IoT) and wearable sensors in the healthcare domain offers a smart way of identifying the symptoms of PD patients. In which smart sensors are worn on the patient’s body which continuously monitor the symptoms in patients and track their possible health status. The major objective of this work is to propose a machine learning-based healthcare model that best classifies the subjects into healthy and Parkinson's patients by extracting the most important features. A step regression-based feature selection method is followed to improve the classification of PD. A Shapiro Wilk test is adopted to check the normality of the gait dataset. This model is implemented on three publicly available Parkinson’s datasets collected from three different studies available on Psyionet. All these data sets contain VGRF recordings obtained from eight different sensors placed under each foot. Experimentation is done on the Jupyter notebook by utilizing Python as a programming language. Experimental results revealed that our proposed model with effective pre-processing, feature extraction, and feature selection method achieved the highest accuracy result of 95.54%, 98.80%, and 94.52% respectively when applied to three datasets. Our research inducts knowledge about significant characteristics of a patient suffering from PD and may help to diagnose and cure at an early stage.

Groundwater Level Prediction Using a Multiple Objective Genetic Algorithm-Grey Relational Analysis Based Weighted Ensemble of ANFIS Models

Predicting groundwater levels is critical for ensuring sustainable use of an aquifer’s limited groundwater reserves and developing a useful groundwater abstraction management strategy. The purpose of this study was to assess the predictive accuracy and estimation capability of various models based on the Adaptive Neuro Fuzzy Inference System (ANFIS). These models included Differential Evolution-ANFIS (DE-ANFIS), Particle Swarm Optimization-ANFIS (PSO-ANFIS), and traditional Hybrid Algorithm tuned ANFIS (HA-ANFIS) for the one- and multi-week forward forecast of groundwater levels at three observation wells. Model-independent partial autocorrelation functions followed by frequentist lasso regression-based feature selection approaches were used to recognize appropriate input variables for the prediction models. The performances of the ANFIS models were evaluated using various statistical performance evaluation indexes. The results revealed that the optimized ANFIS models performed equally well in predicting one-week-ahead groundwater levels at the observation wells when a set of various performance evaluation indexes were used. For improving prediction accuracy, a weighted-average ensemble of ANFIS models was proposed, in which weights for the individual ANFIS models were calculated using a Multiple Objective Genetic Algorithm (MOGA). The MOGA accounts for a set of benefits (higher values indicate better model performance) and cost (smaller values indicate better model performance) performance indexes calculated on the test dataset. Grey relational analysis was used to select the best solution from a set of feasible solutions produced by a MOGA. A MOGA-based individual model ranking revealed the superiority of DE-ANFIS (weight = 0.827), HA-ANFIS (weight = 0.524), and HA-ANFIS (weight = 0.697) at observation wells GT8194046, GT8194048, and GT8194049, respectively. Shannon’s entropy-based decision theory was utilized to rank the ensemble and individual ANFIS models using a set of performance indexes. The ranking result indicated that the ensemble model outperformed all individual models at all observation wells (ranking value = 0.987, 0.985, and 0.995 at observation wells GT8194046, GT8194048, and GT8194049, respectively). The worst performers were PSO-ANFIS (ranking value = 0.845), PSO-ANFIS (ranking value = 0.819), and DE-ANFIS (ranking value = 0.900) at observation wells GT8194046, GT8194048, and GT8194049, respectively. The generalization capability of the proposed ensemble modelling approach was evaluated for forecasting 2-, 4-, 6-, and 8-weeks ahead groundwater levels using data from GT8194046. The evaluation results confirmed the useability of the ensemble modelling for forecasting groundwater levels at higher forecasting horizons. The study demonstrated that the ensemble approach may be successfully used to predict multi-week-ahead groundwater levels, utilizing previous lagged groundwater levels as inputs.

A generalized l 2,p-norm regression based feature selection algorithm

Feature selection is an important data dimension reduction method, and it has been used widely in applications involving high-dimensional data such as genetic data analysis and image processing. In order to achieve robust feature selection, the latest works apply the or -norm of matrix to the loss function and regularization terms in regression, and have achieved encouraging results. However, these existing works rigidly set the matrix norms used in the loss function and the regularization terms to the same or -norm, which limit their applications. In addition, the algorithms for solutions they present either have high computational complexity and are not suitable for large data sets, or cannot provide satisfying performance due to the approximate calculation. To address these problems, we present a generalized -norm regression based feature selection (-RFS) method based on a new optimization criterion. The criterion extends the optimization criterion of (-RFS) when the loss function and the regularization terms in regression use different matrix norms. We cast the new optimization criterion in a regression framework without regularization. In this framework, the new optimization criterion can be solved using an iterative re-weighted least squares (IRLS) procedure in which the least squares problem can be solved efficiently by using the least square QR decomposition (LSQR) algorithm. We have conducted extensive experiments to evaluate the proposed algorithm on various well-known data sets of both gene expression and image data sets, and compare it with other related feature selection methods.

A novel permission-based Android malware detection system using feature selection based on linear regression

With the developments in mobile and wireless technology, mobile devices have become an important part of our lives. While Android is the leading operating system in market share, it is the platform most targeted by attackers. Although many solutions have been proposed in the literature for the detection of Android malware, there is still a need for attribute selection methods to be used in Android malware detection systems. In this study, a machine learning-based malware detection system is proposed to distinguish Android malware from benign applications. At the feature selection stage of the proposed malware detection system, it is aimed to remove unnecessary features by using a linear regression-based feature selection approach. In this way, the dimension of the feature vector is reduced, the training time is decreased, and the classification model can be used in real-time malware detection systems. When the results of the study are examined, the highest 0.961 is obtained according to the F-measure metric by using at least 27 features.

Logistic Regression based Feature Selection and Two-Stage Detection for EEG based Motor Imagery Classification

Electroencephalogram (EEG) based motor imagery (MI) classification requires efficient feature extraction and consistent accuracy for reliable brain-computer interface (BCI) systems. Achieving consistent accuracy in EEGMI classification is still big challenge according to the nature of EEG signal which is subject dependent. To address this problem, we propose a feature selection scheme based on Logistic Regression (LRFS) and two-stage detection (TSD) in channel instantiation approach. In TSD scheme, Linear Discriminant Analysis was utilized in first-stage detection; while Gradient Boosted Tree and k-Nearest Neighbor in second-stage detection. To evaluate the proposed method, two publicly available datasets, BCI competition III-Dataset IVa and BCI competition IV-Dataset 2a, were used. Experimental results show that the proposed method yielded excellent accuracy for both datasets with 95.21% and 94.83%, respectively. These results indicated that the proposed method has consistent accuracy and is promising for reliable BCI systems.

An improved clear cell renal cell carcinoma stage prediction model based on gene sets

BackgroundClear cell renal cell carcinoma (ccRCC) is the most common subtype of renal cell carcinoma and accounts for cancer-related deaths. Survival rates are very low when the tumor is discovered in the late-stage. Thus, developing an efficient strategy to stratify patients by the stage of the cancer and inner mechanisms that drive the development and progression of cancers is critical in early prevention and treatment.ResultsIn this study, we developed new strategies to extract important gene features and trained machine learning-based classifiers to predict stages of ccRCC samples. The novelty of our approach is that (i) We improved the feature preprocessing procedure by binning and coding, and increased the stability of data and robustness of the classification model. (ii) We proposed a joint gene selection algorithm by combining the Fast-Correlation-Based Filter (FCBF) search with the information value, the linear correlation coefficient, and variance inflation factor, and removed irrelevant/redundant features. Then the logistic regression-based feature selection method was used to determine influencing factors. (iii) Classification models were developed using machine learning algorithms. This method is evaluated on RNA expression value of clear cell renal cell carcinoma derived from The Cancer Genome Atlas (TCGA). The results showed that the result on the testing set (accuracy of 81.15% and AUC 0.86) outperformed state-of-the-art models (accuracy of 72.64% and AUC 0.81) and a gene set FJL-set was developed, which contained 23 genes, far less than 64. Furthermore, a gene function analysis was used to explore molecular mechanisms that might affect cancer development.ConclusionsThe results suggested that our model can extract more prognostic information, and is worthy of further investigation and validation in order to understand the progression mechanism.

Cancer-Related Gene Signature Selection Based on Boosted Regression for Multilayer Perceptron

Gene expression profiling is a useful technique for analyzing cellular function, and gene expression profiles are widely studied in human cancer research. Gene expression data usually consist of a very large number of features and a relatively small number of samples, and extracting a small number of important features from these data is a major challenge of gene expression-based analysis in cancer research. In this paper, we propose an embedded feature selection algorithm using boosted linear regression-based feature selection. The boosting technique is applied to derive the ensemble feature selector and improve the performance of linear regression-based feature selection. The proposed feature selection algorithm, called boosted regression-based feature selection for the multilayer perceptron (BREG-MLP), repeats the boosted feature selection process to extract the smallest feature subset while maintaining good classification performance. We apply the proposed BREG-MLP to some human cancer-related gene expression data sets for the purpose of extracting important features, and we confirm that BREG-MLP offers improved performance compared to single regression-based feature selection methods.

Efficient Feature Selection via $\ell _{2,0}$ ℓ2,0-norm Constrained Sparse Regression

Sparse regression based feature selection method has been extensively investigated these years. However, because it has a non-convex constraint, i.e., $\ell _{2,0}$ℓ2,0-norm constraint, this problem is very hard to solve. In this paper, unlike most of the other methods which only solve its slack version by introducing sparsity regularization into objective function forcibly, a novel framework is proposed by us to solve the original $\ell _{2,0}$ℓ2,0-norm constrained sparse regression based feature selection problem. We transform our objective function into Linear Discriminant Analysis (LDA) by using a new label coding method, thus enabling our model to calculate the ratio of inter-class scatter to intra-class scatter of features which is the most widely used feature discrimination evaluation metric. According to that ratio, features can be selected by a simple sorting method. The projection gradient descent algorithm is introduced to further improve the performance of our algorithm by using the solution obtained before as its initial solution. This ensures the stability of this iterative algorithm. We prove that the proposed method can get the global optimal solution of this non-convex problem when all features are statistically independent. For the general case where features are statistically dependent, extensive experiments on six small sample size datasets and one large-scale dataset show that our algorithm has comparable or better classification capability comparing with other eight state-of-the-art feature selection methods by the SVM classifier. We also show that our algorithm can obtain a low loss value, which means the solution of our algorithm can get very close to this NP-hard problem’s real solution. What is more, because we solve the original $\ell _{2,0}$ℓ2,0-norm constrained problem, we avoid the heavy work of tuning the regularization parameter because its meaning is explicit in our method, i.e., the number of selected features. At last, we evaluate the stability of our algorithm from two perspectives, i.e., the objective function values and the selected features, by experiments. From both perspectives, our algorithm shows satisfactory stability performance.

Sparse structural feature selection for multitarget regression

Feature selection is one of the most important dimensionality reduction techniques for its efficiency and interpretation. Recently, some sparse regression-based feature selection methods have obtained an increased attention from the research community. However, previous sparse regression-based feature selection methods are limited by the double-layer structure. To improve the learning performance, in this work, we propose a sparse structural feature selection model for multitarget regression, which utilizes a structure matrix to expand the double-layer to multi-layer structure. Our aim tries to explore the essential inter-target correlations. To enhance the robustness of our proposed method, we emphasize a joint ℓ2, 1-norm minimization on the loss function, regression matrix, and structure matrix. An effective optimization method with provable convergence behavior is also proposed. Extensive experimental results on multivariate prediction demonstrate the effectiveness of our proposed method.

PV power forecasting using an integrated GA-PSO-ANFIS approach and Gaussian process regression based feature selection strategy

This paper presents a hybrid approach for the forecasting of electricity production in microgrids with solar photovoltaic (PV) installations. An accurate PV power generation forecasting tool essentially addresses the issues resulting from the intermittent and uncertain nature of solar power to ensure efficient and reliable system operation. A day-ahead, hourly mean PV power generation forecasting method based on a combination of genetic algorithm (GA), particle swarm optimization (PSO) and adaptive neuro-fuzzy inference systems (ANFIS) is presented in this study. Binary GA with Gaussian process regression model based fitness function is used to determine important input parameters that significantly influence the amount of output power of a PV generation plant; and an integrated hybrid algorithm combining GA and PSO is used to optimize an ANFIS based PV power forecasting model for the plant. The proposed modeling technique is tested based on power generation data obtained from Goldwind microgrid system found in Beijing. Forecasting results demonstrate the superior performance of the proposed method as compared with commonly used forecasting approaches. The proposed approach outperformed existing artificial neural network (ANN), linear regression (LR), and persistence based forecasting models, validating its effectiveness.

Incorporating association rule networks in feature category-weighted naive Bayes model to support weaning decision making

Mechanical ventilation is an invasive intervention commonly used in the intensive care unit to assist patients' respirations. Physicians' decisions to wean patients from ventilation are critical: Effective weaning decisions improve patient care and well-being, but ineffective decisions can create serious severe consequences and complications. Data-driven approaches, enabled by appropriate data mining techniques, can support physicians' weaning decisions. A review of the existing techniques reveals several gaps. Specifically, most techniques assume that a feature can contribute equally to different outcome classes, overlook the “fuzzy region” issue, and assess the importance of individual features holistically rather than scrutinize the discriminant power of distinctive categories of a feature toward each decision outcome class. To address these backdrops, we propose an association rule network-based feature category-weighted naive Bayes method capable of dealing with the inherent challenges in weaning decision making. Our method analyzes feature category weights for each decision outcome by incorporating association rule learning with weighted network analysis, then applies a category-weighted naive Bayes model that can assign differential weights to various feature categories. The results of our empirical evaluation, including several prevalent techniques—artificial neural network (ANN), ANN with backward feature selection, support vector machine (SVM), and SVM with logistical regression based feature selection—indicate that the proposed method consistently outperforms all the benchmark techniques in terms of accuracy, precision, recall and F measure.

Constructing Multi-frequency High-Order Functional Connectivity Network for Diagnosis of Mild Cognitive Impairment.

Human brain functional connectivity (FC) networks, estimated based on resting-state functional magnetic resonance imaging (rs-fMRI), has become a promising tool for imaging-based brain disease diagnosis. Conventional low-order FC network (LON) usually characterizes pairwise temporal correlation of rs-fMRI signals between any pair of brain regions. Meanwhile, high-order FC network (HON) has provided an alternative brain network modeling strategy, characterizing more complex interactions among low-order FC sub-networks that involve multiple brain regions. However, both LON and HON are usually constructed within a fixed and relatively wide frequency band, which may fail in capturing (sensitive) frequency-specific FC changes caused by pathological attacks. To address this issue, we propose a novel "multi-frequency HON construction" method. Specifically, we construct not only multiple frequency-specific HONs (intra-spectrum HONs), but also a series of cross-frequency interaction-based HONs (inter-spectrum HONs) based on the low-order FC sub-networks constructed at different frequency bands. Both types of these HONs, together with the frequency-specific LONs, are used for the complex network analysis-based feature extraction, followed by sparse regression-based feature selection and the classification between mild cognitive impairment (MCI) patients and normal aging subjects using a support vector machine. Compared with the previous methods, our proposed method achieves the best diagnosis accuracy in early diagnosis of Alzheimer's disease.

Integrative approach for inference of gene regulatory networks using lasso-based random featuring and application to psychiatric disorders.

BackgroundInferring gene regulatory networks is one of the most interesting research areas in the systems biology. Many inference methods have been developed by using a variety of computational models and approaches. However, there are two issues to solve. First, depending on the structural or computational model of inference method, the results tend to be inconsistent due to innately different advantages and limitations of the methods. Therefore the combination of dissimilar approaches is demanded as an alternative way in order to overcome the limitations of standalone methods through complementary integration. Second, sparse linear regression that is penalized by the regularization parameter (lasso) and bootstrapping-based sparse linear regression methods were suggested in state of the art methods for network inference but they are not effective for a small sample size data and also a true regulator could be missed if the target gene is strongly affected by an indirect regulator with high correlation or another true regulator.ResultsWe present two novel network inference methods based on the integration of three different criteria, (i) z-score to measure the variation of gene expression from knockout data, (ii) mutual information for the dependency between two genes, and (iii) linear regression-based feature selection.Based on these criterion, we propose a lasso-based random feature selection algorithm (LARF) to achieve better performance overcoming the limitations of bootstrapping as mentioned above.ConclusionsIn this work, there are three main contributions. First, our z score-based method to measure gene expression variations from knockout data is more effective than similar criteria of related works. Second, we confirmed that the true regulator selection can be effectively improved by LARF. Lastly, we verified that an integrative approach can clearly outperform a single method when two different methods are effectively jointed. In the experiments, our methods were validated by outperforming the state of the art methods on DREAM challenge data, and then LARF was applied to inferences of gene regulatory network associated with psychiatric disorders.

Regression-Based Feature Selection on Large Scale Human Activity Recognition

In this paper, we present an approach for regression-based feature selection in human activity recognition. Due to high dimensional features in human activity recognition, the model may have over-fitting and can’t learn parameters well. Moreover, the features are redundant or irrelevant. The goal is to select important discriminating features to recognize the human activities in videos. R-Squared regression criterion can identify the best features based on the ability of a feature to explain the variations in the target class. The features are significantly reduced, nearly by 99.33%, resulting in better classification accuracy. Support Vector Machine with a linear kernel is used to classify the activities. The experiments are tested on UCF50 dataset. The results show that the proposed model significantly outperforms state-of-the-art methods.

Determining Cutoff Point of Ensemble Trees Based on Sample Size in Predicting Clinical Dose with DNA Microarray Data.

Background/Aim. Evaluating the success of dose prediction based on genetic or clinical data has substantially advanced recently. The aim of this study is to predict various clinical dose values from DNA gene expression datasets using data mining techniques. Materials and Methods. Eleven real gene expression datasets containing dose values were included. First, important genes for dose prediction were selected using iterative sure independence screening. Then, the performances of regression trees (RTs), support vector regression (SVR), RT bagging, SVR bagging, and RT boosting were examined. Results. The results demonstrated that a regression-based feature selection method substantially reduced the number of irrelevant genes from raw datasets. Overall, the best prediction performance in nine of 11 datasets was achieved using SVR; the second most accurate performance was provided using a gradient-boosting machine (GBM). Conclusion. Analysis of various dose values based on microarray gene expression data identified common genes found in our study and the referenced studies. According to our findings, SVR and GBM can be good predictors of dose-gene datasets. Another result of the study was to identify the sample size of n = 25 as a cutoff point for RT bagging to outperform a single RT.

Thermal and hyperspectral imaging for Norway spruce (Picea abies) seeds screening

The quality of seeds used in agriculture and forestry is tightly linked to the plant productivity. Thus, the development of high-throughput nondestructive methods to classify the seeds is of prime interest. Visible and near infrared (VNIR, 400–1000nm range) and short-wave infrared (SWIR, 1000–2500nm range) hyperspectral imaging techniques were compared to an infrared lifetime imaging technique to evaluate Norway spruce (Picea abies (L.) Karst.) seed quality. Hyperspectral image and thermal data from 1606 seeds were used to identify viable seeds, empty seeds and seeds infested by Megastigmus sp. larvae. The spectra of seeds obtained from hyperspectral imaging, especially in SWIR range and the thermal signal decay of seeds following an exposure to a short light pulse were characteristic of the seed status. Classification of the seeds to three classes was performed with a Support Vector Machine (nu-SVM) and sparse logistic regression based feature selection. Leave-One-Out classification resulted to 99% accuracy using either thermal or spectral measurements compared to radiography classification. In spectral imaging case, all important features were located in the SWIR range. Furthermore, the classification results showed that accurate (93.8%) seed sorting can be achieved with a simpler method based on information from only three hyperspectral bands at 1310nm, 1710nm and 1985nm locations, suggesting a possibility to build an inexpensive screening device. The results indicate that combined classification methods with hyperspectral imaging technique and infrared lifetime imaging technique constitute practically high performance fast and non-destructive techniques for high-throughput seed screening.

Ensemble model of non-linear feature selection-based Extreme Learning Machine for improved natural gas reservoir characterization

The deluge of multi-dimensional data acquired from advanced data acquisition tools requires sophisticated algorithms to extract useful knowledge from such data. Traditionally, petroleum and natural gas engineers rely on “rules-of-thumb” in the selection of optimal features with much disregard to the hidden patterns in operational data. The traditional multivariate method of feature selection has become grossly inadequate as it is incapable of handling the non-linearity embedded in such natural phenomena. With the application of computational intelligence and its hybrid techniques in the petroleum industry, much improvement has been made. However, they are still incapable of handling more than one hypothesis at a time. Ensemble learning offers robust methodologies to handle the uncertainties in most complex industrial problems. This learning paradigm has not been well embraced in petroleum reservoir characterization despite the persistent quest for increased prediction accuracy. This paper proposes a novel ensemble model of Extreme Learning Machine (ELM) in the prediction of reservoir properties while utilizing the non-linear approximation capability of Functional Networks to select the optimal input features. Different instances of ELM were fed with features selected from different bootstrap samplings of the real-life field datasets. When benchmarked against existing techniques, our proposed ensemble model outperformed the multivariate regression-based feature selection, the conventional bagging and the Random Forest methods with higher correlation coefficient and lower prediction errors. This work confirms the huge potential in the capability of the new ensemble modeling paradigm to improve the prediction of reservoir properties.