Classification Prediction of Breast Cancer Based on Machine Learning

Area of Science:

• Oncology
• Computer Science
• Medical Informatics

Background:

• Breast cancer is a leading cause of cancer-related mortality worldwide.
• Early diagnosis is crucial for improving patient outcomes and survival rates.
• Machine learning offers promising tools for enhancing diagnostic accuracy.

Purpose of the Study:

• To develop and compare machine learning models for breast cancer classification and early diagnosis.
• To identify the most effective model for predicting malignant cancer cells.
• To provide a reference for clinical decision-making in breast cancer detection.

Main Methods:

• Standardization of data to mitigate dimensional influences.
• Feature selection using Pearson correlation test, identifying 15 key features.
• Implementation of XGBoost, random forest, logistic regression, and K-nearest neighbor algorithms.
• Utilization of hierarchical sampling for imbalanced datasets and cross-validation for model optimization.

Main Results:

• The XGBoost model, with an 8:2 train-test split, achieved the highest performance metrics.
• Achieved recall of 1.00, precision of 0.960, accuracy of 0.974, and F1-score of 0.980.
• Model performance varied with different dataset divisions (8:2 vs. 7:3).

Conclusions:

• The XGBoost model shows significant potential for accurate and reliable early breast cancer detection.
• The study highlights the importance of feature selection, data preprocessing, and appropriate evaluation metrics (recall, precision, accuracy, F1-score) in machine learning for medical diagnosis.
• The findings support the use of machine learning as a valuable tool in the fight against breast cancer.