A Comparative Analysis of Breast Cancer Detection and Diagnosis Using Data Visualization and Machine Learning...

Area of Science:

• Oncology
• Data Science
• Medical Informatics

Background:

• Cancer remains a significant global health challenge, particularly in developing nations.
• Breast cancer is a prevalent type, where early and accurate diagnosis is critical for effective treatment.
• Existing research explores various methods for predicting breast tumor types.

Purpose of the Study:

• To conduct a comparative analysis of data visualization and machine learning techniques for breast cancer detection and diagnosis.
• To evaluate the diagnostic performance of multiple machine learning algorithms on breast tumor data.
• To identify the most effective methods for improving breast cancer diagnosis accuracy.

Main Methods:

• Utilized a dataset of breast cancer tumors from the University of Wisconsin Hospital.
• Applied data visualization techniques and machine learning algorithms: logistic regression, k-nearest neighbors, support vector machine, naïve Bayes, decision tree, random forest, and rotation forest.
• Employed R, Minitab, and Python for analysis and visualization.

Main Results:

• Diagnostic performances across the evaluated machine learning applications were comparable.
• The logistic regression model, utilizing all features, achieved the highest classification accuracy at 98.1%.
• The proposed approaches demonstrated enhanced accuracy in breast cancer detection.

Conclusions:

• Data visualization and machine learning techniques offer significant benefits for cancer detection and decision-making processes.
• The study highlights the potential of these computational methods to open new avenues in breast cancer diagnosis.
• Accurate diagnostic tools are crucial for improving patient outcomes in breast cancer treatment.