TY - JOUR AU - Shinta Putri AU - Mustakim Mustakim PY - 2022/06/30 Y2 - 2024/03/28 TI - Dimensional Data Unsupervised Learning Using an Analytic Hierarchy Process in Determining Attributes in the Classification Algorithm JF - Building of Informatics, Technology and Science (BITS) JA - bits VL - 4 IS - 1 SE - Articles DO - 10.47065/bits.v4i1.1752 UR - https://ejurnal.seminar-id.com/index.php/bits/article/view/1752 AB - Systematic keyword is needed in improving the quality of higher education, one of which is the needed to increase the competence of graduates every year. In increasing student graduation, it is necessary to classify student graduation to find out whether the student is said to be on time (TW) or possibility on time (KTW) using the BPNN and PNN methods. The data used is the Alumni data of the 2013-2020 Information System  study program with 7 criteria use, namely GPA, Total Credits, Number of Repetitive Courses, Taking TA Curse in Semester 7, Procrastination, Self-Confidence, and Discipline. The data obtained is then carried out in the process of sharing training data and testing data using K-Means Clustering with the aim; of getting the best accuracy results. Furthermore, the classification stage using BPNN and PNN resulted in an accuracy of 98% and 95% with learning rate of 0.125 and a spread value of 0.1 ER -