fieldid E-Thesis & Research สถาบันเทคโนโลยีไทย-ญี่ปุ่น
สืบค้น:

เขตข้อมูล ข้อมูล
บทคัดย่อ
Performance Comparisons of Machine Learning Approaches for Numerically Structured Breast Cancer Data Classifications : This thesis has compared the performance of a machine learning approaches for numerically structured breast cancer data classifications. The data classification is one of data analytics methods which common use to encourage the decision-making aim of many businesses, medical, healthcare or any requirement that impact to the human life. Furthermore, this thesis has been perceived the motivation from the big data analytics technologies which can drive to improve Thailand technologies in the present trend. Subsequently, the breast cancer dataset was selected from the UCI data repository in the large of a dataset stock to support many researchers to perform the data analytics purposes. Moreover, this dataset contains 30 features of three breast cell nucleus which shown the 569 instances of historical diagnostic. Nonetheless, the machine learning techniques which selected to perform in this thesis can be divided into four algorithms of the Decision Tree (DT), Naive Bayes (NB), Artificial Neural Network (ANN), and Support Vector Machine (SVM). In addition, the cross-validation method is conducted to evaluate the performance of each technique by random the parameter customization to find the highest performance which separated into three categories of accuracy rate, error rate, and classification lead time. The research objective is to find the appropriate of a classifier that has provided the highest performance using the RapidMiner studio 7.4 program. The results have shown that t
ผู้แต่ง
Nattaphon Talmongkol
ประเภทสิ่งพิมพ์
วิทยานิพนธ์ (MET) 2018
เลขหน้า
107
หัวเรื่อง
Engineering Technology
หัวเรื่อง
Engineering Technology--Thesis
หัวเรื่อง
Thai-Nichi Institute of Technology.Engineering Technology--Thesis
เอกสารฉบับเต็ม
Download full text