Article Per Year (5 Year)
p-Index From 2020 - 2025
0.61
P-Index
This Author published in this journals
All Journal ILKOM Jurnal Ilmiah Jurnal Nasional Komputasi dan Teknologi Informasi Indonesian Journal of Intelligence Data Science (IJIDS)
Luther Alexander Latumakulita
Universitas Sam Ratulangi
Aerospace Engineering Automotive Engineering Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering Engineering Library & Information Science Mathematics Neuroscience Transportation

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

Found 3 Documents
Search

 1 
Decision Tree C4.5 Performance Improvement using Synthetic Minority Oversampling Technique (SMOTE) and K-Nearest Neighbor for Debtor Eligibility Evaluation Edi Priyanto; Enny Itje Sela; Luther Alexander Latumakulita; Noourul Islam
ILKOM Jurnal Ilmiah Vol 15, No 2 (2023)
Publisher : Prodi Teknik Informatika FIK Universitas Muslim Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33096/ilkom.v15i2.1676.373-381

Abstract

Nowadays, information technology especially machine learning has been used to evaluate the feasibility of debtors. One of the challenges in this classification model is the occurrence of imbalanced datasets, especially in the German Credit Dataset. Another challenge is developing an optimal model for evaluating debtor eligibility. Based on these challenges, this study aims to develop an optimal model for evaluating debtor eligibility on the German Credit Dataset, using the decision trees, k-Nearest Neighbor (k-NN) and Synthetic Minority Oversampling Technique (SMOTE). SMOTE and k-NN is used to overcome challenges regarding imbalanced datasets. While the decision tree are applied to produce a debtor classification model. In general, the steps taken are preparing datasets, pre-processing data, dividing datasets, oversampling with SMOTE, and classification models using decision trees, and testing. Model performance evaluation is represented by accuracy values obtained from the confusion matrix and area under curve (AUC) values generated by the Receiver Operating Characteristic (ROC). Based on the tests that have been carried out, the best accuracy value in the test is obtained at 73.00% and the AUC value is 0.708, in parameters k = 3 and Max-Depth = 25. Based on the analysis produced, the proposed model can improve performance compared to if the dataset is not applied SMOTE.
Perbandingan Tiga Skema Kombinasi Hyper Parameter Convolutional Neural Networks dalam Klasifikasi Biji Kopi Hasil Roasting Luther Alexander Latumakulita
Jurnal Nasional Komputasi dan Teknologi Informasi (JNKTI) Vol 7, No 3 (2024): Juni 2024
Publisher : Program Studi Teknik Komputer, Fakultas Teknik. Universitas Serambi Mekkah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32672/jnkti.v7i3.7580

Abstract

Abstrak - Indonesia merupakan negara kelima dengan konsumen kopi tertinggi di dunia. Selain enak, kopi juga bermanfaat untuk meningkatkan metabolisme tubuh. Nilai konsumen kopi yang bertambah harus dibarengi dengan peningkatan mutu hasil roasting biji kopi. Tujuan dari penelitian ini yaitu untuk membangun sistem kecerdasan buatan yang dapat melakukan klasifikasi biji kopi robusta setelah proses roasting menggunakkan algoritma Convolutional Neural Network (CNN). Sebanyak 450 data biji kopi dari 3 class klasifikasi dilatih menggunakan 3 skema dengan kombinasi nilai epoch, batch size, dan learning rate yang berbeda. Hasil ekseprimen menunjukan bahwa skema dengan kombinasi hyper parameter dengan nilai epoch sebesar 200, batch size sebesar 16, dan learning rate sebesar 0,0001 menghasilkan akurasi testing tertinggi 96% dibandikan dengan kedua skema lainnya yang menghasilkan akurasi testing berturut-turut sebesar 95% dan 93%. Model klasifikasi yang dihasilkan menunjukan performansi system yang sangat baik menandakan model yang ditemukam dalam research ini dapat dipakai untuk mensortir kualitas biji kopi hasil roasting sehingga dapat berdampak positip dalam memajukan industry rosting kopi Indonesia..Kata kunci: CNN, Deep Learning, Klasifikasi, Biji Kopi Robusta Abstract - Indonesia is the fifth country with the highest coffee consumers in the world. Apart from being delicious, coffee is also useful for increasing the body's metabolism. The increasing consumer value of coffee must be accompanied by an increase in the quality of roasted coffee beans. The aim of this research is to build an artificial intelligence system that can classify robusta coffee beans after the roasting process using the Convolutional Neural Network (CNN) algorithm. A total of 450 coffee bean data from 3 classification classes were trained using 3 schemes with different combinations of epoch, batch size and learning rate values. The experimental results show that the scheme with a combination of hyper parameters with an epoch value of 200, a batch size of 16, and a learning rate of 0.0001 produces the highest testing accuracy of 96% compared to the other two schemes which produce testing accuracy of 95% and 93% respectively.  The resulting classification model shows very good system performance, indicating that the model found in this research can be used to sort the quality of roasted coffee beans so that it can have a positive impact in advancing the Indonesian coffee rosting industry.Keywords : CNN, Deep Learning, Classification, Robusta Coffee Beans
Sistem Pakar Diagnosa Penyakit Lambung Menggunakan Metode Forward Chaining Dan Certainty Factor Scheryl Pongantung; Marline Sofiana Paendong; Luther Alexander Latumakulita
Indonesian Journal of Intelligence Data Science Vol 3 No 2 (2024): Volume 3 No 2 2024
Publisher : Faculty of Mathematics and Natural Sciences Sam Ratulangi University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35799/ijids.v3i2.50076

Abstract

Limited knowledge about the early symptoms of stomach diseases has motivated the author to develop a system that helps the community obtain information. This system aims to provide assistance to the public in obtaining information, consultation, and early treatment for stomach diseases without having to have direct meetings with experts. The expertise of a medical professional in diagnosing stomach diseases can be implemented into an application. In this Expert System, Forward Chaining method is used for reasoning and the Certainty Factor method is used to calculate confidence levels. Based on data processing from one of the users, the research results show that GERD is the most likely diagnosis, with a Certainty Factor value of 96.5%.
Co-Authors Edi Priyanto Enny Itje Sela Marline Sofiana Paendong Noourul Islam Scheryl Pongantung