Rizky Tri Asmono
Universitas Dian Nuswantoro

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 3 Documents
Search

Absolute Correlation Weighted Naïve Bayes for Software Defect Prediction Asmono, Rizky Tri; Wahono, Romi Satria; Syukur, Abdul
Journal of Software Engineering Vol 1, No 1 (2015)
Publisher : IlmuKomputer.Com

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (652.927 KB)

Abstract

The maintenance phase of the software project can be very expensive for the developer team and harmful to the users because some flawed software modules. It can be avoided by detecting defects as early as possible. Software defect prediction will provide an opportunity for the developer team to test modules or files that have a high probability defect. Naïve Bayes has been used to predict software defects. However, Naive Bayes assumes all attributes are equally important and are not related each other while, in fact, this assumption is not true in many cases. Absolute value of correlation coefficient has been proposed as weighting method to overcome Naïve Bayes assumptions. In this study, Absolute Correlation Weighted Naïve Bayes have been proposed. The results of parametric test on experiment results show that the proposed method improve the performance of Naïve Bayes for classifying defect-prone on software defect prediction.
PENERAPAN GINI INDEX DAN K-NEAREST NEIGHBOR UNTUK KLASIFIKASI TINGKAT KOGNITIF SOAL PADA TAKSONOMI BLOOM Setiyorini, Tyas; Asmono, Rizky Tri
Jurnal Pilar Nusa Mandiri Vol 13 No 2 (2017): Pilar Nusa Mandiri : Journal of Computing and Information System Periode Septemb
Publisher : LPPM Universitas Nusa Mandiri

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (893.144 KB)

Abstract

Sebagai pedoman dalam merancang ujian yang layak, yang terdiri dari soal-soal yang memiliki berbagai tingkatan secara kognitif, Taksonomi Bloom telah diterapkan secara luas. Saat ini, kalangan pendidik mengidentifikasi tingkat kognitif soal pada Taksonomi Bloom masih menggunakan cara manual. Hanya sedikit pendidik yang dapat mengidentifikasi tingkat kognitif dengan benar, sebagian besar melakukan kesalahan dalam mengklasifikasikan soal-soal. K-Nearest Neighbor (KNN) adalah metode yang efektif untuk klasifikasi tingkat kognitif soal pada Taksonomi Bloom, tetapi KNN memiliki kelemahan yaitu kompleksitas komputasi kemiripan datanya besar apabila dimensi fitur datanya tinggi. Untuk menyelesaikan kelemahan tersebut diperlukan metode Gini Index untuk mengurangi dimensi fitur yang tinggi. Beberapa percobaan dilakukan untuk memperoleh arsitektur yang terbaik dan menghasilkan klasifikasi yang akurat. Hasil dari 10 percobaan pada dataset Question Bank dengan KNN diperoleh akurasi tertinggi yaitu 59,97% dan kappa tertinggi yaitu 0,496. Kemudian pada KNN+Gini Index diperoleh akurasi tertinggi yaitu 66,18% dan kappa tertinggi yaitu 0,574. Berdasarkan hasil tersebut maka dapat disimpulkan bahwa Gini Index mampu mengurangi dimensi fitur yang tinggi, sehingga meningkatkan kinerja KNN dan meningkatkan tingkat akurasi klasifikasi tingkat kognitif soal pada Taksonomi Bloom.
IMPLEMENTATION OF GAIN RATIO AND K-NEAREST NEIGHBOR FOR CLASSIFICATION OF STUDENT PERFORMANCE Setiyorini, Tyas; Asmono, Rizky Tri
Jurnal Pilar Nusa Mandiri Vol 16 No 1 (2020): Pilar Nusa Mandiri : Journal of Computing and Information System Publishing Peri
Publisher : LPPM Universitas Nusa Mandiri

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (895.904 KB) | DOI: 10.33480/pilar.v16i1.813

Abstract

Predicting student performance is very useful in analyzing weak students and providing support to students who face difficulties. However, the work done by educators has not been effective enough in identifying factors that affect student performance. The main predictor factor is an informative student academic score, but that alone is not good enough in predicting student performance. Educators utilize Educational Data Mining (EDM) to predict student performance. KK-Nearest Neighbor is often used in classifying student performance because of its simplicity, but the K-Nearest Neighbor has a weakness in terms of the high dimensional features. To overcome these weaknesses, a Gain Ratio is used to reduce the high dimension of features. The experiment has been carried out 10 times with the value of k is 1 to 10 using the student performance dataset. The results of these experiments are obtained an average accuracy of 74.068 with the K-Nearest Neighbor and obtained an average accuracy of 75.105 with the Gain Ratio and K-Nearest Neighbor. The experimental results show that Gain Ratio is able to reduce the high dimensions of features that are a weakness of K-Nearest Neighbor, so the implementation of Gain Ratio and K-Nearest Neighbor can increase the accuracy of the classification of student performance compared to using the K-Nearest Neighbor alone.