ENHANCING DETECTION METHOD OF BREAST CANCER USING COIMBRA DATASET

  • ADIL HANI ABDULKAREEM İstanbul Ticaret Üniversitesi
  • Mustafa Cem Kasapbaşı
Keywords: Coimbra dataset, Cancer biomarker, Machine learning

Abstract

Breast cancer is one of the most dangerous and second most common types of cancer in the world. Breast cancer-fighting with developed devices and medical therapies has become easier. To obtain the best result in breast cancer treatment, periodic checks should be carried out to follow the early diagnosis. Data Mining techniques are used to predict the success of treatment or diagnosis. In this study, the K-Nearest Neighbor (k-NN), Naïve Bayes classifier algorithms of machine learning were used for early detection of breast cancer. From the UC Irvine Machine Learning Repository (UCI) library Coimbra Breast Cancer data set which consists of age, glucose, body mass index (BMI), resistin, insulin, adiponectin, homeostatic model assessment (HOMA), monocyte chemoattractant protein-1 (MCP1), and leptin attributes were used. K-NN model using Age, Resistin, Glucose,  and BMI give the highest results, where 90% of specificity 84% percent of sensitivity, and 87.5% accuracy is achieved. These findings provide promising evidence that models combining resistin, glucose, age, and BMI may be a powerful tool for breast cancer detection.

Published
2020-08-30
How to Cite
[1]
A. ABDULKAREEM and M. Kasapbaşı, “ENHANCING DETECTION METHOD OF BREAST CANCER USING COIMBRA DATASET”, jtas, vol. 3, no. 1, pp. 51-59, Aug. 2020.