Recursive Feature Elimination Research Articles

Estimating Leaf Area Index in Apple Orchard by UAV Multispectral Images with Spectral and Texture Information

The Leaf Area Index (LAI) strongly influences vegetation evapotranspiration and photosynthesis rates. Timely and accurately estimating the LAI is crucial for monitoring vegetation growth. The unmanned aerial vehicle (UAV) multispectral digital camera platform has been proven to be an effective tool for this purpose. Currently, most remote sensing estimations of LAIs focus on cereal crops, with limited research on economic crops such as apples. In this study, a method for estimating the LAI of an apple orchard by extracting spectral and texture information from UAV multispectral images was proposed. Specifically, field measurements were conducted to collect LAI data for 108 sample points during the final flowering (FF), fruit setting (FS), and fruit expansion (FE) stages of apple growth in 2023. Concurrently, UAV multispectral images were obtained to extract spectral and texture information (Gabor transform). The Support Vector Regression Recursive Feature Elimination (SVR-REF) was employed to select optimal features as inputs for constructing models to estimate the LAI. Finally, the optimal model was used for LAI mapping. The results indicate that integrating spectral and texture information effectively enhances the accuracy of LAI estimation, with the relative prediction deviation (RPD) for all models being greater than 2. The Categorical Boosting (CatBoost) model established for FF exhibits the highest accuracy, with a validation set R2, root mean square error (RMSE), and RPD of 0.867, 0.203, and 2.482, respectively. UAV multispectral imagery proves to be valuable in estimating apple orchard LAIs, offering real-time monitoring of apple growth and providing a scientific basis for orchard management.

Advancing emotion recognition in facial expressions through PCA, RFE, and MLP Integration

Emotion recognition through facial expressions is vital for enhancing human-computer interaction, making systems more intuitive and responsive to user needs. This study introduces an innovative approach to emotion detection, leveraging Principal Component Analysis (PCA) and Recursive Feature Elimination (RFE) for feature extraction and optimization. The methodology focuses on refining facial feature representations to improve classification accuracy, which is critical for accurately detecting emotions like happiness, sadness, fear, and surprise. The approach was applied to two widely recognized facial emotion datasets: the Cohn-Kanade (CK) and the Japanese Female Facial Expression (JAFFE) datasets. By integrating PCA and RFE, the model efficiently selects the most relevant features, enhancing the overall performance of emotion recognition. A comparative analysis with existing deep learning models highlights the advantages of the proposed method. The effectiveness of this approach is further supported by the results, where the model achieves an accuracy of 98.49% on the CK dataset and 96.03% on the JAFFE dataset. These results demonstrate a significant improvement over recent methods, indicating the model's potential for real-world applications.

GSAAN with War Strategy Optimization Algorithm for Cyber Security in a 5G Wireless Communication Network

In this paper, the graph sample and aggregate-attention network with war strategy optimization algorithm for cyber security in the 5G wireless communication network (CS-5GWCN-GSAAN-WSOA) is proposed in 5G mobile networks to identify cyber threats. Initially, the input data are amassed from the 5G-NIDD dataset. Then the input data are fed to preprocessing, here redundancy elimination and missing value replacement are performed utilizing the contrast limited adaptive histogram equalization filtering (CLAHEF) method. After preprocessing, optimal features are selected by the stochastic gradient boosting based recursive feature elimination process. The selected features are fed to the graph sample and aggregate-attention network (GSAAN) to detect cyber-attacks in 5G wireless communication. Generally, the GSAAN approach does not adopt any optimization techniques for determining optimal parameters and guaranteeing accurate attack detection. Therefore, the war strategy optimization algorithm (WSOA) contemplates to optimize the GSAAN weight parameters. The CS-5GWCN-GSAAN-WSOA method is activated on Python. The CS-5GWCN-GSAAN-WSOA method attains 22.51%, 20.35%, and 35.54% higher accuracy, 19.45%, 27.80%, and 18.93% higher sensitivity analysed with existing models, like cyber security attack identification in 5G wireless systems utilizing machine learning (CS-5GWCN-SVM), enhanced dropping attacks detection in 5G networks depending on deep learning models (CS-5GWCN-KNN), and cyber security issues in 5G enabled attack detection through deep learning (CS-5GWCN-RNN), respectively.

Predicting the efficiency of chidamide in patients with angioimmunoblastic T-cell lymphoma using machine learning algorithm.

Chidamide is subtype-selective histone deacetylase (HDAC) inhibitor that showed promising result in clinical trials to improve prognosis of angioimmunoblastic T-cell lymphoma (AITL) patients. However, in real world settings, contradictory reports existed as to whether chidamide improve overall survival (OS). Therefore, we aimed to develop an interpretable machine learning (Machine learning)-based model to predict the 2-year overall survival of AITL patients based on chidamide usage and baseline features. A total of 183 patients with AITL were randomly divided into training set and testing set. We used 5ML algorithms to build predictive models. Recursive feature elimination (RFE) method was used to filter for the most important features. The ML models were interpreted and the relevance of the selected features was determined using the Shapley additive explanations (SHAP) method and the local interpretable model-agnostic explanationalgorithm. A total of 183 patients with newly diagnosed AITL from 2012 to 2022 from 3 centers in China were enrolled in our study. Seventy-one patients were dead within 2years after diagnosis. Five ML algorithms were built based on chidamide usage and 16 baseline features to predict 2-year OS. Catboost model presented to be the best predictive model. After RFE screening, 12 variables demonstrated the best performance (AUC = 0.8651). Using chidamide ranked third among all the variables that correlated with 2-year OS. This study demonstrated that the Catboost model with 12 variables could effectively predict the 2-year OS of AITL patients. Combining chidamide in the treatment therapy was positively correlated with longer OS of AITL patients.

Machine Learning Based Abnormal Gait Classification with IMU Considering Joint Impairment.

Gait analysis systems are critical for assessing motor function in rehabilitation and elderly care. This study aimed to develop and optimize an abnormal gait classification algorithm considering joint impairments using inertial measurement units (IMUs) and walkway systems. Ten healthy male participants simulated normal walking, walking with knee impairment, and walking with ankle impairment under three conditions: without joint braces, with a knee brace, and with an ankle brace. Based on these simulated gaits, we developed classification models: distinguishing abnormal gait due to joint impairments, identifying specific joint disorders, and a combined model for both tasks. Recursive Feature Elimination with Cross-Validation (RFECV) was used for feature extraction, and models were fine-tuned using support vector machine (SVM), random forest (RF), and extreme gradient boosting (XGB). The IMU-based system achieved over 91% accuracy in classifying the three types of gait. In contrast, the walkway system achieved less than 77% accuracy in classifying the three types of gait, primarily due to high misclassification rates between knee and ankle joint impairments. The IMU-based system shows promise for accurate gait assessment in patients with joint impairments, suggesting future research for clinical application improvements in rehabilitation and patient management.

A Highland Barley Crop Extraction Method Based on Optimized Feature Combination of Multiple Phenological Sentinel-2 Images

Previous studies have primarily focused on the extraction of highland barley crops using single phenological images, which ignored the selection of the optimal phenological period for classification. Utilizing the multiple phenological images from Sentinel-2 to construct 25 features, including spectral, red edge, vegetation, and texture features, the recursive feature elimination algorithm and the random forest algorithm (RF) were employed to optimize feature datasets for different phenological stages, which were then used for the identification and classification of high-land barley by RF. The main results were as follows: (1) Information extraction based on feature optimization combinations yielded good overall classification accuracy, with classification accuracies for highland barley being 92.56% (jointing stage), 90.90% (heading stage), 90.74% (flowering stage), 91.55% (milk ripening stage), and 90.51% (maturity stage), respectively. (2) NDVIre1 had the highest importance score (0.1792) in the feature selection combination, indicating that the red edge index contributed significantly to crop information extraction and classification. (3) The five feature variables—GLCM_Mean, RVI, homogeneity, MAX, and GLCM_Correlation—showed stability and universality in the extraction of highland barley. These results demonstrated that the images that derived from the jointing and milk ripening phenological stages had the best applicability for highland barley extraction, and the optimized feature datasets that composed of NDVIre1 were conductive to detect and monitor of highland barley crops in the mountainous regions of northwest China.

Artificial neural network prediction of postoperative complications in papillary thyroid microcarcinoma based on preoperative ultrasonographic features.

To predict post-thyroidectomy complications in papillary thyroid microcarcinoma (PTMC) patients using a deep learning model based on preoperative ultrasonographic features. This study addresses the global rise in PTMC incidence and the challenges in treatment decision-making with high-resolution ultrasonography. This study enrolled 1638 patients with clinically staged cN0 PTMC who received surgical treatment from 1997 to 2019 at Beijing Friendship Hospital. Deep learning model was developed using fully connected neural network. Feature selection included 1000 iterations of Bootstrap sampling and Recursive Feature Elimination (RFE) to identify the top 10 features. Data preprocessing involved normalization and imputation for missing values. SMOTE addressed class imbalance. The model was trained and tested on random data split, with performance metrics including Accuracy (ACC), Area Under the Curve (AUC), Sensitivity (SEN), and Specificity (SPE), visualized through a ROC curve and confusion matrix. The fully connected deep neural network model demonstrated high accuracy (ACC 0.81), Area Under the Curve (AUC 0.74), sensitivity (SEN 0.65), and specificity (SPE 0.83) and visualized by ROC curve and confusion matrix. These results highlight the model's reliability and potential as an effective tool in predicting postoperative complications and assisting in clinical decision-making for PTMC patients. This study highlights the potential of deep learning in enhancing medical predictions and personalized healthcare. Despite promising results, limitations include a single-center data source and unconsidered factors like lifestyle and genetics. Future research should expand data sources, include more influencing factors, and refine algorithms to improve accuracy and applicability in thyroid cancer treatment.

Performance prediction of sludge volume index of oxygenic photogranule based wastewater treatment system using machine learning algorithms

Oxygenic Photogranulation is a novel biotechnology that treats wastewater without external aeration and produces biomass in dense photogranules with high settling velocities. Oxygenic photogranules (OPG) based wastewater treatment (WWT) faces challenges during scaleup due to its dynamic and complex system variables, making troubleshooting costly. The Machine Learning (ML) approach can address this issue by creating a WWT process simulation. Moreover, traditional mechanistic models do not capture the interaction between input and output features due to their high dimensionality and the non-linear relationship, making them computationally expensive. In this study, the two-stage feature selection (FS) method is studied to enhance the prediction performance of SVI30, a critical operational parameter, to ensure optimal settleability and minimal loss of active photogranules. The optimal feature subsets generated by the two-stage selection method were evaluated using four regression models: decision tree, random forest, gradient boosting, and extreme gradient boosting. Results indicate that, among all regression models, the decision tree performs well having a prediction efficiency of 85 % with the subset of features obtained after Recursive Feature Elimination (RFE) of decision tree features in the second stage. This indicates the effectiveness of the two-stage FS method in identifying the most relevant features for predicting SVI30. The structured approach of FS and model evaluation highlights the potential of ML in addressing complex operational challenges in OPG WWT operations.

Cognitive frailty in relation to vitamin B12 and 25-hydroxyvitamin D in an elderly population: a cross-sectional study from NHANES.

Nutritional support has been identified as a potential intervention for cognitive frailty; however, the association between 25-hydroxyvitamin D [25-(OH)D], vitamin B12, and cognitive frailty remains ambiguous. This study utilized data from two cycles (2011-2012, 2013-2014) of the National Health and Nutrition Examination Survey (NHANES) to investigate this relationship. The researchers constructed a 41-item frailty index encompassing diverse aspects of physical functioning, psychological evaluation, and medical conditions, and evaluated each participant individually. The study utilized Spearman's rank correlation coefficient and univariate ordered logistic regression to assess the relationships between variables and cognitive frailty. Recursive feature elimination and cross-validation methods were employed to identify the most influential variables for building and optimizing multivariate ordered logistic regression models. Subgroup analyses and interaction tests were further conducted to validate the identified correlations. The findings of this study confirm a negative linear correlation between 25-(OH)D levels and cognitive frailty in older adults. Specifically, a one-unit increase in 25-(OH)D levels was associated with a 12% reduction in the risk of cognitive frailty. The result was further supported by subgroup analyses and interaction tests. The existence of a negatively correlated linear association between 25-(OH)D levels and cognitive frailty in older adults is plausible, but further rigorously designed longitudinal studies are necessary to validate this relationship.

Mental issues, internet addiction and quality of life predict burnout among Hungarian teachers: a machine learning analysis

BackgroundBurnout is usually defined as a state of emotional, physical, and mental exhaustion that affects people in various professions (e.g. physicians, nurses, teachers). The consequences of burnout involve decreased motivation, productivity, and overall diminished well-being. The machine learning-based prediction of burnout has therefore become the focus of recent research. In this study, the aim was to detect burnout using machine learning and to identify its most important predictors in a sample of Hungarian high-school teachers.MethodsThe final sample consisted of 1,576 high-school teachers (522 male), who completed a survey including various sociodemographic and health-related questions and psychological questionnaires. Specifically, depression, insomnia, internet habits (e.g. when and why one uses the internet) and problematic internet usage were among the most important predictors tested in this study. Supervised classification algorithms were trained to detect burnout assessed by two well-known burnout questionnaires. Feature selection was conducted using recursive feature elimination. Hyperparameters were tuned via grid search with 10-fold cross-validation. Due to class imbalance, class weights (i.e. cost-sensitive learning), downsampling and a hybrid method (SMOTE-ENN) were applied in separate analyses. The final model evaluation was carried out on a previously unseen holdout test sample.ResultsBurnout was detected in 19.7% of the teachers included in the final dataset. The best predictive performance on the holdout test sample was achieved by random forest with class weigths (AUC = 0.811; balanced accuracy = 0.745, sensitivity = 0.765; specificity = 0.726). The best predictors of burnout were Beck’s Depression Inventory scores, Athen’s Insomnia Scale scores, subscales of the Problematic Internet Use Questionnaire and self-reported current health status.ConclusionsThe performances of the algorithms were comparable with previous studies; however, it is important to note that we tested our models on previously unseen holdout samples suggesting higher levels of generalizability. Another remarkable finding is that besides depression and insomnia, other variables such as problematic internet use and time spent online also turned out to be important predictors of burnout.

Leveraging SEER data through machine learning to predict distant lymph node metastasis and prognosticate outcomes in hepatocellular carcinoma patients.

This study aims to develop and validate machine learning-based diagnostic and prognostic models to predict the risk of distant lymph node metastases (DLNM) in patients with hepatocellular carcinoma (HCC) and to evaluate the prognosis for this cohort. Utilizing a retrospective design, this investigation leverages data extracted from the Surveillance, Epidemiology, and End Results (SEER) database, specifically the January 2024 subset, to conduct the analysis. The study cohort consists of 15,775 patients diagnosed with HCC as identified within the SEER database, spanning 2016 to 2020. In the construction of the diagnostic model, recursive feature elimination (RFE) is employed for variable selection, incorporating five critical predictors: age, tumor size, radiation therapy, T-stage, and serum alpha-fetoprotein (AFP) levels. These variables are the foundation for a stacking ensemble model, which is further elucidated through Shapley Additive Explanations (SHAP). Conversely, the prognostic model is crafted utilizing stepwise backward regression to select pertinent variables, including chemotherapy, radiation therapy, tumor size, and age. This model culminates in the development of a prognostic nomogram, underpinned by the Cox proportional hazards model. The outcome of the diagnostic model is the occurrence of DLNM in patients. The outcome of the prognosis model is determined by survival time and survival status. The integrated model developed based on stacking demonstrates good predictive performance and high interpretative variability and differentiation. The area under the curve (AUC) in the training set is 0.767, while the AUC in the validation set is 0.768. The nomogram, constructed using the Cox model, also demonstrates consistent and strong predictive capabilities. At the same time, we recognized elements that have a substantial impact on DLNM and the prognosis and extensively discussed their significance in the model and clinical practice. Our study identified key predictive factors for DLNM and elucidated significant prognostic indicators for HCC patients with DLNM. These findings provide clinicians with valuable tools to accurately identify high-risk individuals for DLNM and conduct more precise risk stratification for this patient subgroup, potentially improving management strategies and patient outcomes.

Determination of Possible Biomarkers for Predicting Well-Differentiated Thyroid Cancer Recurrence by Different Ensemble Machine Learning Methods

Objective: Well-differentiated thyroid cancer (WDTC) is the most common thyroid malignancy and although it is curable, the risk of recurrence is high. In this study, classification algorithms based on clinicopathologic features of WDTC patients were used to determine the possible of recurrence in WDTC and to evaluate potential predictive factors, and possible biomarkers based on the optimal model were identified. Method: In this study, open access data on 383 patients with WDTC, 108 with recurrence and 275 without recurrence, were used. In order to predict recurrence in WDTC patients, features were selected using recursive feature elimination variable selection method among features and classification was performed with two ensemble learning methods (Random Forest, Adaboost). Results: Two different ensemble learning models used to classify recurrence in WDTC were Random Forest with an accuracy of 0.957, sensitivity of 0.889, specificity of 0.978, positive predictive value of 0.923, negative predictive value of 0.967, Matthews correlation coefficient of 0.878, G-mean of 0.945, F1-score of 0.906, and accuracy of 0.940, sensitivity of 0.889, specificity of 0.955, positive predictive value of 0.857, negative predictive value of 0.966, Matthews correlation coefficient of 0.833, G-mean of 0.910, F1-score of 0.873. Conclusion: According to variable importance based on the Random Forest, the 5 possible clinical biomarkers for predicting WDTC recurrence are Response, Risk, Node, Tumor, and age. In the light of these findings, patient management and treatment planning can be organized.

CatBoost Optimization Using Recursive Feature Elimination

CatBoost Optimization Using Recursive Feature Elimination

Deep Learning-Based Prediction of Daily COVID-19 Cases Using X (Twitter) Data.

Due to the importance of COVID-19 control, innovative methods for predicting cases using social network data are increasingly under attention. This study aims to predict confirmed COVID-19 cases using X (Twitter) social network data (tweets) and deep learning methods. We prepare data extracted from tweets by natural language processing (NLP) and consider the daily G-value (growth rate) as the target variable of COVID-19, collected from the worldometer. We develop and evaluate a time series mixer (TSMixer) predictive model for multivariate time series. The mean squared error (MSE) loss on the test dataset was 0.0063 for 24-month Gvalue prediction when using the MinMax normalization with recursive feature elimination (RFE) and average or min aggregation method. Our findings illuminate the potential of integrating social media data to enhance daily COVID-19 case predictions and are applicable also for epidemiological forecasting purposes.

SRFE: A stepwise recursive feature elimination approach for network intrusion detection systems

SRFE: A stepwise recursive feature elimination approach for network intrusion detection systems

An accident severity prediction framework with consider features interaction

Accident severity prediction is increasingly important for preventing and reducing traffic accident losses. However, many studies always ignore the complex relationship between features during feature selection. To improve the prediction accuracy of the accident severity prediction model, this paper considers the interactions between multiple features. First, the feature selection algorithm of Recursive Feature Elimination with Cross-Validation (RFECV) is improved by using Shapley Additive explanations (SHAP) as the feature importance assessment metric. Then, to decrease the time expense of manually finding hyperparameters of the model, the Hunter Prey Optimization (HPO) algorithm is introduced and logistic mapping together with stochastic perturbation is added to it, which makes it easier to skip out of the partial optimum during the optimization search. Finally, the improved HPO algorithm is used to optimize the hyperparameters of the CatBoost model. The US traffic accident dataset is introduced for the validity of the proposed framework. Experimental results show that the proposed framework achieves a prediction accuracy of 96.63%, which is better than other state-of-the-art methods. The high accuracy of the prediction model can help decision-makers develop more rational transportation policies, and this study also proposes some traffic management measures based on the selected features.

Prediction accuracy and repeatability of UAV based biomass estimation in wheat variety trials as affected by variable type, modelling strategy and sampling location

BackgroundThis study explores the use of Unmanned Aerial Vehicles (UAVs) for estimating wheat biomass, focusing on the impact of phenotyping and analytical protocols in the context of late-stage variety selection programs. It emphasizes the importance of variable selection, model specificity, and sampling location within the experimental plot in predicting biomass, aiming to refine UAV-based estimation techniques for enhanced selection accuracy and throughput in variety testing programs.ResultsThe research uncovered that integrating geometric and spectral traits led to an increase in prediction accuracy, whilst a recursive feature elimination (RFE) based variable selection workflowled to slight reductions in accuracy with the benefit of increased interpretability. Models, tailored to specific experiments were more accurate than those modelling all experiments together, while models trained for broad-growth stages did not significantly increase accuracy. The comparison between a permanent and a precise region of interest (ROI) within the plot showed negligible differences in biomass prediction accuracy, indicating the robustness of the approach across different sampling locations within the plot. Significant differences in the within-season repeatability (w2) of biomass predictions across different experiments highlighted the need for further investigation into the optimal timing of measurement for prediction.ConclusionsThe study highlights the promising potential of UAV technology in biomass prediction for wheat at a small plot scale. It suggests that the accuracy of biomass predictions can be significantly improved through optimizing analytical and modelling protocols (i.e., variable selection, algorithm selection, stage-specific model development). Future work should focus on exploring the applicability of these findings under a wider variety of conditions and from a more diverse set of genotypes.

Improved snow depth estimation on the Tibetan Plateau using AMSR2 and ensemble learning models

Snow depth (SD) is essential for studying climate change and hydrological cycle on the Tibetan Plateau (TP). Despite the effectiveness of passive microwave remote sensing for large-scale SD measurement, its low spatial resolution and scanning gaps limit its application, particularly in the TP region where the terrain is complex and snow distribution exhibits obvious heterogeneity. This study developed Advanced Microwave Scanning Radiometer 2 (AMSR2) SD downscaling models for the TP using ensemble learning methods and AMSR2 brightness temperature data from October 1, 2012, to April 30, 2021. We employed five ensemble methods—AdaBoost, GBDT, XGBoost, LightGBM, and Random Forest—with LightGBM achieving the highest accuracy (RMSE=2.66 cm). Recursive feature elimination (RFE) was applied to the LightGBM model, optimizing factor selection and maintaining high accuracy. The models excelled in estimating shallow snow areas (SD<5 cm) with an RMSE of 1.60 cm. SHapley Additive exPlanations (SHAP) values were used to quantify global and local contributions of each factor in the modeling process. Key factors included snow cover days, meteorological influences, and brightness temperature (BT) at 89 GHz with horizontal polarization, although their contributions varied significantly across the TP due to environmental gradients. The resulting 500 m SD estimates offer detailed and accurate snow distribution information in complex mountainous regions. Our results help to improve water resource management and climate change analysis on the TP.

Prediction of 28-Day All-Cause Mortality in Heart Failure Patients with Clostridioides difficile Infection Using Machine Learning Models: Evidence from the MIMIC-IV Database.

Heart failure (HF) may induce bowel hypoperfusion, leading to hypoxia of the villa of the bowel wall and the occurrence of Clostridioides difficile infection (CDI). However, the risk factors for the development of CDI in HF patients have yet to be fully illustrated, especially because of a lack of evidence from real-world data. Clinical data and survival situations of HF patients with CDI admitted to ICU were extracted from the Medical Information Mart for Intensive Care (MIMIC)-IV database. For developing a model that can predict 28-day all-cause mortality in HF patients with CDI, the Recursive Feature Elimination with Cross-Validation (RFE-CV) method was used for feature selection. And nine machine learning (ML) algorithms, including logistic regression (LR), decision tree, Bayesian, adaptive boosting, random forest (RF), gradient boosting decision tree, XGBoost, light gradient boosting machine, and categorical boosting, were applied for model construction. After training and hyperparameter optimization of the models through grid search 5-fold cross-validation, the performance of models was evaluated by the area under curve (AUC), accuracy, sensitivity, specificity, precision, negative predictive value, and F1 score. Furthermore, the SHapley Additive exPlanations (SHAP) method was used to interpret the optimal model. A total of 526 HF patients with CDI were included in the study, of whom 99 cases (18.8%) experienced death within 28 days. Eighteen of the 57 variables were selected for the model construction algorithm for model construction. Among the ML models considered, the RF model emerged as the optimal model achieving the accuracy, F1-score, and AUC values of 0.821, 0.596, and 0.864, respectively. The net benefit of the model surpassed other models at 16%-22% threshold probabilities based on decision curve analysis. According to the importance of features in the RF model, red blood cell distribution width, blood urea nitrogen, Simplified Acute Physiology Score II, Sequential Organ Failure Assessment, and white blood cell count were highlighted as the five most influential variables. We developed ML models to predict 28-day all-cause mortality in HF patients associated with CDI in the ICU, which are more effective than the conventional LR model. The RF model has the best performance among all the ML models employed. It may be useful to help clinicians identify high-risk HF patients with CDI.

Machine learning models for differential diagnosing HER2-low breast cancer: A radiomics approach.

To develop machine learning models based on preoperative dynamic enhanced magnetic resonance imaging (DCE-MRI) radiomics and to explore their potential prognostic value in the differential diagnosis of human epidermal growth factor receptor 2 (HER2)-low from HER2-positive breast cancer (BC). A total of 233 patients with pathologically confirmed invasive breast cancer admitted to our hospital between January 2018 and December 2022 were included in this retrospective analysis. Of these, 103 cases were diagnosed as HER2-positive and 130 cases were HER2 low-expression BC. The Synthetic Minority Oversampling Technique is employed to address the class imbalance problem. Patients were randomly split into a training set (163 cases) and a validation set (70 cases) in a 7:3 ratio. Radiomics features from DCE-MRI second-phase imaging were extracted. Z-score normalization was used to standardize the radiomics features, and Pearson's correlation coefficient and recursive feature elimination were used to explore the significant features. Prediction models were constructed using 6 machine learning algorithms: logistic regression, random forest, support vector machine, AdaBoost, decision tree, and auto-encoder. Receiver operating characteristic curves were constructed, and predictive models were evaluated according to the area under the curve (AUC), accuracy, sensitivity, and specificity. In the training set, the AUC, accuracy, sensitivity, and specificity of all models were 1.000. However, in the validation set, the auto-encoder model's AUC, accuracy, sensitivity, and specificity were 0.994, 0.976, 0.972, and 0.978, respectively. The remaining models' AUC, accuracy, sensitivity, and specificity were 1.000. The DeLong test showed no statistically significant differences between the machine learning models in the training and validation sets (Z = 0, P = 1). Our study investigated the feasibility of using DCE-MRI-based radiomics features to predict HER2-low BC. Certain radiomics features showed associations with HER2-low BC and may have predictive value. Machine learning prediction models developed using these radiomics features could be beneficial for distinguishing between HER2-low and HER2-positive BC. These noninvasive preoperative models have the potential to assist in clinical decision-making for HER2-low breast cancer, thereby advancing personalized clinical precision.