A Novel Double Machine Learning Approach for Detecting Early Breast Cancer Using Advanced Feature Selection and Dimensionality Reduction Techniques

Sep 24, 2026

Paper proposes three Double Machine Learning (DML) models to enhance early breast cancer detection accuracy using machine learning on a breast cancer dataset. Each DML model learns primary structured and non-linear features, fuses learned representations via a meta-classifier, and outputs final predictions. The first fuses Random Forest interpretability with a feedforward neural network, the second combines XGBoost with an ANN, and the third integrates LightGBM with an ANN for sequential dependencies. Combined with PCA and feature selection, the system reaches accuracy up to 0.99.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages23

Reading Time58 min

AccessFree to download

Introduction

Breast cancer background and motivation

Proposed DML framework

Model 1: Random Forest + Feedforward Neural Network

Model 2: XGBoost + Artificial Neural Network

Model 3: LightGBM + Artificial Neural Network

Feature selection and dimensionality reduction

PCA and improved classification performance

[www. nature.com/scientificreports](www. nature.com/scientificreports) OPEN A novel double machine learning approach for detecting early breast cancer using advanced featureselection and dimensionality reduction techniques Suganya Athisayamani1, Tamilazhagan S2, A. Robert Singh3, Jae-Yong Hwang4􀀍 & Gyanendra Prasad Joshi5􀀍 In this paper, three Double Machine Learning (DML) models are proposed to enhance the accuracy of breast cancer detection using machine learning techniques using breast cancer detection dataset. The DML models learn the primary features using machine learning and deep learning models. Then, these features are fused by a meta-classifier to achieve the best classification performance. The first DML model combines the interpretability of Random Forest (RF) with the deep learning capabilities of a Feedforward Neural Network (FNN). RF processes structured features, providing class probabilities and feature importance scores, while the FNN learns non-linear relationships and generates embeddings. These outputs are fused into a combined feature vector, which is then used by a meta-classifier for final predictions. This approach effectively captures both structured features and non-linear patterns, making it suitable for datasets with complex dependencies. The second model pairs eXtreme Gradient Boosting (XGBoost), a highly efficient boosting algorithm for tabular data, with an Artificial Neural Network (ANN) . XGBoost optimizes decision tree ensembles and provides class probabilities, while the ANN processes numerical data to learn deeper representations. A meta-classifier then uses the fused outputs from both XGBoost and ANN for final predictions. This model is particularly effective for datasets combining structured features (handled by XGBoost) with numerical features (handled by ANN). The third model integrates LightGBM, a fast and scalable gradient-boosting framework, with an ANN, which is well-suited for analyzing sequential data. LightGBM processes structured features to provide probabilities and importance scores, while the ANN learns temporal dependencies from sequential data. The outputs from LightGBM and ANN are concatenated and passed into a metaclassifier for decision-making. This model is ideal for datasets with both static features (LightGBM) and continuous data (ANN), such as time-series datasets or datasets with sequential dependencies. These DML models, when combined with dimensionality reduction (PCA) and feature selection, significantly improve the performance of breast cancer detection systems by leveraging both structured and sequential data with high accuracy of 0.99. Keywords Breast Cancer, Machine Learning, SVM, Random Forest, Decision Tree, KNN, PCA, Feature Selection, Double machine learning Machine Learning and Deep learning models1 are used in various applications to automate the tasks. These models are capable of processing numerical data2, text data, images3, videos and time series data. Cancer is a disease caused by the uncontrollable growth of body cells4. The Human Body is made up of trillions of cells. Uncontrollable growth in a part of the body can also affect other parts because cancer cells grow faster. The types 1Department of Computing Technologies, SRM Institute of Science and Technology, Kattankulathur, Tamil Nadu, India. 2School of Computing, Sastra Deemed to be University, Thanjavur, Tamil Nadu, India. 3Department of Computational Intelligence, SRM Institute of Science and Technology, Kattankulathur, Tamil Nadu, India. 4Department of Information & Communication Engineering, Daejeon University, 34520 Daejeon, Republic of Korea. 5Department of AI Software, Kangwon National University, 25913 Samcheok, Republic of Korea. 􀀍 email: [platans@dju.kr](platans@dju.kr); [joshi@kangwon.ac.kr](joshi@kangwon.ac.kr) [www. nature.com/scientificreports/](www. nature.com/scientificreports/) of cancer are colorectal cancer, breast cancer, melanoma, prostate cancer, pancr

FAQs

What is the main goal of the proposed paper?

To propose and evaluate distinct DML models that improve the accuracy of early breast cancer detection using machine learning techniques.

How do the three DML models differ from each other?

They pair different base learners: (1) Random Forest with a feedforward neural network, (2) XGBoost with an ANN, and (3) LightGBM with an ANN, with each fusion fed to a meta-classifier.

Why are feature selection and PCA used in the approach?

Feature selection and dimensionality reduction help improve breast cancer detection performance by leveraging structured and sequential/non-linear patterns more effectively.

Sep 24, 2026

Uploaded by eBook King