Statistical analysis of random forest classifier in breast cancer diagnosis
|
Full Text |
Pdf
|
|
Author |
Ahmed Fahad
|
|
e-ISSN |
1819-6608 |
|
On Pages
|
177-187
|
|
Volume No. |
20
|
|
Issue No. |
4
|
|
Issue Date |
April 12, 2025
|
|
DOI |
https://doi.org/10.59018/022530
|
|
Keywords |
breast cancer, machine learning, random forest, x-ray mammography, statistical analysis.
|
Abstract
Breast cancer is an aggressive growth of cells that originates in the breast tissue. This kind of growth can potentially extend to other areas of the body. However, traditional methods for detecting breast cancer, such as ultrasound scanning, magnetic resonance imaging, and X-ray mammography, come with drawbacks, including high expenses, harmful radiation exposure, and patient inconvenience. These issues motivated researchers to use alternative methods for the detection of breast cancer, such as machine learning. Moreover, most algorithms presented for the detection of this cancer used one database, had fewer performance measures, and did not include statistical analysis to investigate the performance deeply. Therefore, this paper presents the Random Forest (RF) algorithm as a classifier. This algorithm is trained and tested using two breast cancer databases. The performance of the classifier algorithm concerning the two databases is evaluated using many performance measurements and analysed statistically. The experimental outcomes indicate the promising performance of the algorithm, where this algorithm outperforms its comparatives in the diagnosis of breast cancer.
Back