Blood donation is a process of taking blood from donors that is declared feasible, in terms of various factors including age, weight, blood pressure, hemoglobin levels, and donor status which are taken into consideration during the feasibility test. This study was conducted to find the most appropriate method with high accuracy and Area Under Curve (AUC) values using 3710 blood donor datasets from the Bekasi City PMI, processed using the Naïve Bayes algorithm method, K-Nearest Neighbors and Decision Tree C4.5. The analysis shows that the Decision Tree C4.5 algorithm shows higher accuracy of 93.83% compared to Naïve Bayes algorithm which shows an accuracy value of 85.15% and the K-Nearest Neighbors algorithm with an accuracy value of 84.10%. In addition to these values, Decision Tree C4.5 is also visually superior where the Decision Tree has an output model tree that shows attribute relationships and has an AUC value of 0.978, Naïve Bayes with an AUC value of 0.927 and K-Nearest Neighbors with an AUC value of 0.816.

The purpose of this study is to help small clubs from Italian Serie A in finding the minimum targets to avoid relegation into Serie B competition (below Serie A league). Relegation will reduce the club’s income from TV revenues and the decline of enthusiastic supporters. Based on the data from the final standings (seasons 2006 until 2018), this can be explained by the Decision Tree method using the C4.5 algorithm. The methods used in this study are data collection, data pre-processing, model proposal, model testing, and model validation. In this study, it is expected that the value of accuracy exceeds 85% to achieve a proper classification.

No abstract