OPTIMASI ALGORITMA JARINGAN SYARAF TIRUAN PADA KETEPATAN STUDI MAHASISWA

BACHRI, Otong Saeful and Nurhayati, Oky Dwi and Subagio, Agus (2018) OPTIMASI ALGORITMA JARINGAN SYARAF TIRUAN PADA KETEPATAN STUDI MAHASISWA. Masters thesis, School of Postgraduate.

[img]
Preview
PDF
469Kb
[img]
Preview
PDF
161Kb
[img]
Preview
PDF
398Kb
[img]
Preview
PDF
179Kb
[img]PDF
Restricted to Repository staff only

1559Kb
[img]
Preview
PDF
15Kb
[img]
Preview
PDF
31Kb

Abstract

Masa studi menjadi salah satu indikator refisiensi proses pendidikan di perguruan tinggi, untuk itu prediksi ketepatan studi bagi mahasiswa perlu dilakukan agar masastudikuliah tidak melebihi waktu yang sudah ditentukan. Beberapa variabel yang menentukan lama studi antara lain variable Indeks Prestasi (IP) semester 1dan semester 2, Indeks Prestasi Kumulatif (IPK) semester 6, jumlah mata kuliah yang diambil, jumlah mata kuliah mengulang. Jaringan syaraf tiruan adalah algoritma yang tepat yang digunakan melakukan proses prediksi dengan mengelompokan data kriteria karena memiliki kemampuan update bobot melalui proses pembelajaran untuk mencapai nilai error yang cukup kecil dan hasil yang diperoleh dari optimasi jaringan syaraf tiruan adalah tingkat keakurasian dalam memprediksi hasil ketepatan studi mahasiswa. Data yang dapat digunakan sebagai input Jaringan Syaraf Tiruan (JST) adalah tahun masuk, IP 1, IP 2, dan IP 3.Hasil studi mengungkapkan bahwa sistem yang diperoleh memiliki tingkat klasifikasi yang baik dan mendekati sempurna, karena nilai Precisiondan Recall yang diperoleh sama-sama lebih besar dari 90%, bahkan pada 10-fold Cross Validation, nilai Recall mencapai 98.1%. Semakin besar nilai k, maka nilai Precision, recall, dan accuracy darisistem tersebut cenderung meningkat. Kata Kunci: optimasi, prediksi, masa studi, jaringan syaraf tiruan The period of study becomes one of the indicators of education process reficiency in universities, therefore the prediction of the accuracy of the study for students needs to be conducted so that the study period of the students will not exceed the time specified. In this study, some of the variables that were examined to determine the length of study include the Achievement Index (GPA) from semester 1 to semester 6, the number of taken courses, the number of repeat courses. Hence, artificial neural network is considered as the correct algorithm used to make the prediction process by grouping the data criteria because of its ability to update the weights through the learning process to achieve a fairly small error value. The results obtained from the optimization of artificial neural networks are the level of accuracy in predicting the accuracy of the student study period. Therefore, data that can be used as input of Artificial Neural Network (ANN) is the variable of year of entry, GPA semester 1, 2, and 3. The results of the study reveal that the system has a good and near perfect Classification, since the Precision and Recall values obtained are both greater than 90%, even at 10-fold Cross Validation, and the Recall value reaches 98.1%. Furthermore, the results show that the higher thevalue of k, the higher thePrecision, recall, and accuracy values of the system. Keywords: optimization, prediction, study period, artificial neural network

Item Type:Thesis (Masters)
Subjects:Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions:School of Postgraduate > Master Program in Information System
ID Code:64234
Deposited By:INVALID USER
Deposited On:04 Sep 2018 09:59
Last Modified:04 Sep 2018 09:59

Repository Staff Only: item control page