<![CDATA[One of the main tasks of Point Career Center (PCC) is to record the suitability of alumni's work with the study program. PCC also provides information about job vacancies. If PCC can predict the suitability of jobs that alumni will get based on academic data and non-academic data, it will help PCC to take policies in an effort to increase the percentage of job suitability of alumni. This research was used BPNN to train dataset with 5 atributtes, namely Grade Point Average, high school background, and competency certificate ownership. Non-academic data are parents’s occupation. BPNN is applied to 70:30, 75:25, 80:20, and 90:10 ratio with several learning rates and several hidden units. The results of this research are accuracy, precision, and recall in all scenarios is above 70%, and the best performance is the ratio of 80:20 and 90:10 with accuracy = 83.33%, precision = 87.50%, and recall = 83.33%. That indicates BPNN is suitable to classify suitability between alumni’s occupation and study program.Keywords: Accuracy, BPNN, Classify, Precision, and Recall]]>
Copyrights © 2022
