ExhauFS: exhaustive search-based feature selection for classification and survival regression

Abstract

Motivation: Feature selection is one of the main techniques used to prevent overfitting in machine learning applications. The most straightforward approach for feature selection is exhaustive search: one can go over all possible feature combinations and pick up the model with the highest accuracy. This method together with its optimizations were actively used in biomedical research, however, publicly available implementation is missing. Results: We present ExhauFS - the user-friendly command-line implementation of the exhaustive search approach for classification and survival regression. Aside from tool description, we included three application examples in the manuscript to comprehensively review the implemented function-ality. First, we executed ExhauFS on a toy cervical cancer dataset to illustrate basic concepts. Then, a multi-cohort microarray and RNA-seq breast cancer datasets were used to construct gene signatures for 5-year recurrence classification. Finally, Cox survival regression models were used to fit isomiR signatures for overall survival prediction for patients with colorectal cancer. Availability: Source codes and documentation of ExhauFS are available on GitHub: https://github.com/s-a-nersisyan/ExhauFS.

ECM–Receptor Regulatory Network and Its Prognostic Role in Colorectal Cancer

“…The ExhauFS tool ( Nersisyan et al, 2021c ) was used to fit Cox survival regression models. For each length of prognostic signature ( k = 1, 2, … , 10), we first selected n most individually predictive features (see the next paragraph for the details) and then fit Cox models for all possible feature subsets.…”

“…The ExhauFS tool (Nersisyan et al, 2021c) was used to fit Cox survival regression models. For each length of prognostic signature (k 1, 2, .…”

“…For this analysis, we employed a novel technique of exhaustive search-based Cox model fitting. Namely, ExhauFS software ( Nersisyan et al, 2021c ) was used to construct prognostic models for all gene/5’-isomiR pairs, triples, etc; then, the best performing model was picked.…”

Restaurant survival prediction using customer-generated content: An aspect-based sentiment analysis of online reviews

“…Condition 3.2 specified the same input variables X as Condition 3.1; however, to optimize CSF prediction, we filtered the input variables X and proposed an ABSA-CSF model with feature selection technology, specifically the exhaustive search method. This approach can choose the model with optimal accuracy by reviewing all possible feature combinations (Nersisyan et al, 2022). It thus does not ignore feature sets that might display optimal prediction performance.…”

Breast Cancer Diagnosis By IoB Implanted Tag Design And Machine Learning

“…Most machine learning methods need to normalize the data. [25] . Data normalization is a technique for standardizing the range of characteristics without affecting the data's dimension.…”