Reproducible large-scale inference in high-dimensional nonlinear models

Radiomics is a technique that extracts quantitative features from medical images using data-characterization algorithms. Radiomic features can be used to identify tissue characteristics and radiologic phenotyping that is not observable by clinicians. A typical workflow for a radiomics study includes cohort selection, radiomic feature extraction, feature and predictive model selection, and model training and validation. While there has been increasing attention given to radiomic feature extraction, standardization, and reproducibility, currently, there is a lack of rigorous evaluation of feature selection methods and predictive models. Herein, we review the published radiomics investigations in CT lung cancer and provide an overview of the commonly used radiomic feature selection methods and predictive models. We also compare limitations of various methods in clinical applications and present sources of uncertainty associated with those methods. This review is expected to help raise awareness of the impact of radiomic feature and model selection methods on the integrity of radiomics studies.

With the rapid development of artificial intelligence in recent years, the research on image processing, text mining, and genome informatics has gradually deepened, and the mining of large-scale databases has begun to receive more and more attention. The objects of data mining have also become more complex, and the data dimensions of mining objects have become higher and higher. Compared with the ultra-high data dimensions, the number of samples available for analysis is too small, resulting in the production of high-dimensional small sample data. High-dimensional small sample data will bring serious dimensional disasters to the mining process. Through feature selection, redundancy and noise features in high-dimensional small sample data can be effectively eliminated, avoiding dimensional disasters and improving the actual efficiency of mining algorithms. However, the existing feature selection methods emphasize the classification or clustering performance of the feature selection results and ignore the stability of the feature selection results, which will lead to unstable feature selection results, and it is difficult to obtain real and understandable features. Based on the traditional feature selection method, this paper proposes an ensemble feature selection method, Random Bits Forest Recursive Clustering Eliminate (RBF-RCE) feature selection method, combined with multiple sets of basic classifiers to carry out parallel learning and screen out the best feature classification results, optimizes the classification performance of traditional feature selection methods, and can also improve the stability of feature selection. Then, this paper analyzes the reasons for the instability of feature selection and introduces a feature selection stability measurement method, the Intersection Measurement (IM), to evaluate whether the feature selection process is stable. The effectiveness of the proposed method is verified by experiments on several groups of high-dimensional small sample data sets.

SummaryModel selection is crucial both to high-dimensional learning and to inference for contemporary big data applications in pinpointing the best set of covariates among a sequence of candidate interpretable models. Most existing work implicitly assumes that the models are correctly specified or have fixed dimensionality, yet both model misspecification and high dimensionality are prevalent in practice. In this paper, we exploit the framework of model selection principles under the misspecified generalized linear models presented in Lv & Liu (2014), and investigate the asymptotic expansion of the posterior model probability in the setting of high-dimensional misspecified models. With a natural choice of prior probabilities that encourages interpretability and incorporates the Kullback–Leibler divergence, we suggest using the high-dimensional generalized Bayesian information criterion with prior probability for large-scale model selection with misspecification. Our new information criterion characterizes the impacts of both model misspecification and high dimensionality on model selection. We further establish the consistency of covariance contrast matrix estimation and the model selection consistency of the new information criterion in ultrahigh dimensions under some mild regularity conditions. Our numerical studies demonstrate that the proposed method enjoys improved model selection consistency over its main competitors.

Software defect prediction models are used for predicting high risk software components. Feature selection has significant impact on the prediction performance of the software defect prediction models since redundant and unimportant features make the prediction model more difficult to learn. Ensemble feature selection has recently emerged as a new methodology for enhancing feature selection performance. This paper proposes a new multi-criteria-decision-making (MCDM) based ensemble feature selection (EFS) method. This new method is termed as MCDM-EFS. The proposed method, MCDM-EFS, first generates the decision matrix signifying the feature’s importance score with respect to various existing feature selection methods. Next, the decision matrix is used as the input to well-known MCDM method TOPSIS for assigning a final rank to each feature. The proposed approach is validated by an experimental study for predicting software defects using two classifiers K-nearest neighbor (KNN) and naïve bayes (NB) over five open-source datasets. The predictive performance of the proposed approach is compared with existing feature selection algorithms. Two evaluation metrics – nMCC and G-measure are used to compare predictive performance. The experimental results show that the MCDM-EFS significantly improves the predictive performance of software defect prediction models against other feature selection methods in terms of nMCC as well as G-measure.

QSAR analysis of diaryl COX-2 inhibitors: Comparison of feature selection and train-test data selection methods

The paper proposes a hybrid approach to reduce dimension in text classification problems, to overcome the issue of Curse of Dimensionality. This hybrid approach is a combination of Feature Selection (FS) and Feature Extraction (FE) methods, considering different aspects of feature relevance, to effectively reduce the dimension in large text datasets. It prevents feature selection biased in favor of a particular FS method. Many FS methods like Term Variance, Document Frequency, Information Gain, Shannons Entropy measure, Mean-Median and Mean Absolute Difference, were implemented and a comparative study was made on their performance when implemented in a hybrid system. The features selected by the individual FS methods are merged using three approaches, namely, Union, Intersection and Modified Union. The sub lists of features further undergo Feature Extraction by PCA, and the reduced feature sub list is clustered with k-means. Finally, the sentiment-score of the individual clusters are calculated using SentiWordNet database which gives the polarity of the data. The experiments were conducted on the benchmark datasets namely Reuters-21,578 and Classic4. The performance evaluation of the system made using the measures like precision, recall, f-score and accuracy shows that the proposed method has improved performance compared to its competitive methods.

In many real-world unsupervised learning applications, given data with balanced distribution, that is, there are an approximately equal number of instances in each class, we often need to construct a model to reveal such balance. However, in many data, especially the high-dimensional ones, the data in the original feature space often do not present such balance due to the redundant and noisy features. To tackle this problem, we apply an unsupervised spectral feature selection method to select some informative features, which can better reveal the balanced structure of data. Although spectral feature selection is one of the most popular unsupervised feature selection methods and has been widely studied, none of the existing spectral feature selection methods consider the balance property of data. To address this issue, in this article, we propose a novel balanced spectral feature selection (BSFS) method, which not only selects the discriminative features but also picks those to reveal the balanced structure of data. To the best of our knowledge, this is the first spectral feature selection method considering balance structure of data. By introducing a balanced regularization term, we integrate the balanced spectral clustering and feature selection into a unified framework seamlessly. At last, the experiments on benchmark datasets show that the proposed one outperforms the conventional feature selection methods in both clustering performance and balance, which demonstrates the effectiveness and efficiency of the proposed method.

Modern manufacturing processes are often characterized by high dimensionality and nonlinearity; dimension reduction is usually required to identify the critical features and help improve the process yield. Dimension reduction and analysis of data with high nonlinearity has been a challenging task. Many methods have been proposed based on manifold learning, which aims to learn a lower-dimensional manifold from the original high-dimensional space. However, since the mappings are usually implicit, it is difficult to build a connection between the classification results and the original features. Apart from the prediction results, industry data analysis also focus on the features themselves. Therefore, methods that are able to combine feature selection with dimension reduction is often needed. This paper proposed a hybrid method for nonlinear dimension reduction and feature selection based on Locally Linear Embedding (LLE). LLE and many filter feature selection shares a common procedure of neighbor search. By integrating nearest neighbor search for both LLE and ReliefF, and adjusting the distance measure in supervised LLE with results from feature selection, the proposed method could connect the feature selection process with the nonlinear model, extract the critical features for further analysis, and improve the performance of process modelling. Two dataset from real industry processes are also illustrated as case studies.

Drug-Protein Interaction (DPI) identification is crucial in drug discovery. The high dimensionality of drug and protein features poses challenges for accurate interaction prediction, necessitating the use of computational techniques. Docking-based methods rely on 3D structures, while ligand-based methods have limitations such as reliance on known ligands and neglecting protein structure. Therefore, the preferred approach is the chemogenomics-based approach using machine learning, which considers both drug and protein characteristics for DPI prediction. In machine learning, feature selection plays a vital role in improving model performance, reducing overfitting, enhancing interpretability, and making the learning process more efficient. It helps extract meaningful patterns from drug and protein data while eliminating irrelevant or redundant information, resulting in more effective machine-learning models. On the other hand, classification is of great importance as it enables pattern recognition, decision-making, predictive modeling, anomaly detection, data exploration, and automation. It empowers machines to make accurate predictions and facilitates efficient decision-making in DPI prediction. For this research work, protein data was sourced from the KEGG database, while drug data was obtained from the DrugBank data machine-learning base. To address the issue of imbalanced Drug Protein Pairs (DPP), different balancing techniques like Random Over Sampling (ROS), Synthetic Minority Over-sampling Technique (SMOTE), and Adaptive SMOTE were employed. Given the large number of features associated with drugs and proteins, feature selection becomes necessary. Various feature selection methods were evaluated: Correlation, Information Gain (IG), Chi-Square (CS), and Relief. Multiple classification methods, including Support Vector Machines (SVM), Random Forest (RF), Adaboost, and Logistic Regression (LR), were used to predict DPI. Finally, this research identifies the best balancing, feature selection, and classification methods for accurate DPI prediction. This comprehensive approach aims to overcome the limitations of existing methods and provide more reliable and efficient predictions in drug-protein interaction studies.

• 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.

Target spectrum based feature selection (TSFS): A new method based on chain coding for target detection problems

In the era of digital information, online platforms play a crucial role in shaping public opinion. However, the extensive spread of misinformation and fake news poses a significant challenge, largely fueled by non-credible users. Detecting user credibility is vital for ensuring the reliability of information on these platforms. This study employs supervised machine learning algorithms, leveraging key user features to enhance credibility detection. Feature selection methods, specifically SelectKBest and correlation-based algorithms, are explored for their impact on X-Platform user credibility detection. Utilizing various classifiers, including support vector machine, logistic regression, and XGBoost, experiments are conducted on the ArPFN dataset, which is a labeled, balanced, publicly available dataset. The evaluation includes measures like accuracy, precision, recall, and F1-score to assess efficiency. This research considers feature categories and selection methods with SML to detect their impact on the accuracy of X-Platform user credibility detection, making this research a reference for researchers and practitioners working in the field of SML, feature engineering, and social media analysis. We aim to advance the field’s understanding of effective strategies for mitigating the spread of fake news. The novelty of this study lies in the comprehensive exploration of feature selection methods and their influence on credibility detection, contributing valuable insights for future research in this domain.

Stock market prediction is a very noisy problem and the use of any additional information to increase accuracy is necessary. In this paper, for the stock daily return prediction problem, the set of features is expanded to include indicators not only for the stock to be predicted itself but also a set of other stocks and currencies. Afterwards, different feature selection and classification methods are utilized for prediction. The daily close returns of the 3 most traded stocks (GARAN, THYAO, and ISCTR) in Borsa Istanbul (BIST) are predicted using indicators computed on those stocks, indicators for all the other stocks listed in the BIST100 index, and indicators on the dollar-gold prices. Twenty-five different indicators on daily stock prices are computed to form feature vectors for each trading day. These feature vectors are assigned class labels according to the daily close returns. Expanding the feature space with BIST100 stocks features results in a high dimensional feature space, with possibly noisy or irrelevant features. Therefore, feature selection methods are utilized to select the most informative features. In order to determine relevance scores of features, fast filter-based methods, gain ratio and relief, are used. Experiments are performed based on individual stock features, dollar-gold features (DG), BIST100 stock features (BIST100), and a combination of BIST100 and DG with and without feature selection. Using the gain ratio feature selection with a gradient boosting machine (GBM), the movements of GARAN stock were predicted with an accuracy of 0.599 and an F-measure of 0.614. For THYAO, the relief feature selection with the GBM gave an accuracy of 0.558, and for ISCTR, the gain ratio feature selection with logistic regression achieved an accuracy of 0.581. It was found that using BIST100 stock features boosts classification results for all stocks in terms of accuracy.

<p>Symbolic regression (SR) is a function identification process, the task of which is to identify and express the relationship between the input and output variables in mathematical models. SR is named to emphasise its ability to find the structure and coefficients of the model simultaneously. Genetic Programming (GP) is an attractive and powerful technique for SR, since it does not require any predefined model and has a flexible representation. However, GP based SR generally has a poor generalisation ability which degrades its reliability and hampers its applications to science and real-world modeling. Therefore, this thesis aims to develop new GP approaches to SR that evolve/learn models exhibiting good generalisation ability. This thesis develops a novel feature selection method in GP for high-dimensional SR. Feature selection can potentially contribute not only to improving the efficiency of learning algorithms but also to enhancing the generalisation ability. However, feature selection is seldom considered in GP for high-dimensional SR. The proposed new feature selection method utilises GP’s built-in feature selection ability and relies on permutation to detect the truly relevant features and discard irrelevant/noisy features. The results confirm the superiority of the proposed method over the other examined feature selection methods including random forests and decision trees on identifying the truly relevant features. Further analysis indicates that the models evolved by GP with the proposed feature selection method are more likely to contain only the truly relevant features and have better interpretability. To address the overfitting issue of GP when learning from a relatively small number of instances, this thesis proposes a new GP approach by incorporating structural risk minimisation (SRM), which is a framework to estimate the generalisation performance of models, into GP. The effectiveness of SRM highly depends on the accuracy of the Vapnik-Chervonenkis (VC) dimension measuring model complexity. This thesis significantly extends an experimental method (instead of theoretical estimation) to measure the VC-dimension of a mixture of linear and nonlinear regression models in GP for the first time. The experimental method has been conducted using uniform and non-uniform settings and provides reliable VC-dimension values. The results show that our methods have an impressively better generalisation gain and evolve more compact model, which have a much smaller behavioural difference from the target models than standard GP and GP with bootstrap, The proposed method using the optimised non-uniform setting further improves the one using the uniform setting. This thesis employs geometric semantic GP (GSGP) to tackle the unsatisfied generalisation performance of GP for SR when no overfitting occurs. It proposes three new angle-awareness driven geometric semantic operators (GSO) including selection, crossover and mutation to further explore the geometry of the semantic space to gain a greater generalisation improvement in GP for SR. The angle-awareness brings new geometric properties to these geometric operators, which are expected to provide a greater leverage for approximating the target semantics in each operation, and more importantly, to be resistant to overfitting. The results show that compared with two kinds of state-of-the-art GSOs, the proposed new GSOs not only drive the evolutionary process fitting the target semantics more efficiently but also significantly improve the generalisation performance. A further comparison on the evolved models shows that the new method generally produces simpler models with a much smaller size and containing important building blocks of the target models.</p>

During the drug development process, it is common to carry out toxicity tests and adverse effect studies, which are essential to guarantee patient safety and the success of the research. The use of in silico quantitative structure-activity relationship (QSAR) approaches for this task involves processing a huge amount of data that, in many cases, have an imbalanced distribution of active and inactive samples. This is usually termed the class-imbalance problem and may have a significant negative effect on the performance of the learned models. The performance of feature selection (FS) for QSAR models is usually damaged by the class-imbalance nature of the involved datasets. This paper proposes the use of an FS method focused on dealing with the class-imbalance problems. The method is based on the use of FS ensembles constructed by boosting and using two well-known FS methods, fast clustering-based FS and the fast correlation-based filter. The experimental results demonstrate the efficiency of the proposal in terms of the classification performance compared to standard methods. The proposal can be extended to other FS methods and applied to other problems in cheminformatics.