Diseases Hematology Prediction Using Data Mining Techniques

Abstract

The purpose of this dissertation to study the procedures of data mining methods using methods and techniques in a trained format, which is part of the data mining techniques, which are divided into two sub-sets: The training data set and test data set, the training data set used to create the model while the test data set used to find the effectiveness of the model by evaluating the effectiveness of the model is using a confusion matrix The data in this research are based on the results of blood tests during the period 2019-2020 in the laboratory of Hamesa hospital as data in this research. It uses Data Mining in three ways: Decision Tree, Neural Network, Naive Bayes and Orange Tool to analyze of the research, then take the best model into the prototype in program development.

          From the experimental results so the results can be as follows: From the data set of 1543 people, the risk value is 742 people and the non-risk person is 801 people. In the Decision Tree format with 97.1% accuracy, in the Neural Network with 90.4% and in the Naive Bayes mode with 88.2% accuracy, then take the model the most accurate is: Decision Tree as a model for program development.