Articles
<![CDATA[Pengamanan Data Teks Dengan NTRU dan Modulus Function pada Koefisien IHWT Citra Warna ]]>
<![CDATA[Ronsen Purba]]>;
<![CDATA[Irpan Pardosi]]>;
<![CDATA[Harry Darmawan]]>;
<![CDATA[Aldo Alex Sitorus]]>
<![CDATA[ Jurnal SIFO Mikroskil Vol 20, No 1 (2019): JSM Volume 20 Nomor 1 Tahun 2019 ]]>
Publisher : <![CDATA[Fakultas Informatika ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (541.284 KB)
|
DOI: 10.55601/jsm.v20i1.649
<![CDATA[The development of digital information have caused the rise of information technology security to protect text data that contains secret. Steganography is one of many solutions for securing text data by hiding the text data on an image so that another party would not know the existence of such data. Criteria of a good steganography involves imperceptibility, fidelity, robustness dan recovery. One steganographic method is CD (Coefficient Difference), adopted from PVD (Pixel Value Differencing) which does hiding in spatial domain using difference of two pixel values that results in large modification of pixel values, reducing imperceptibility. Modulus function is used to solve such shortcoming in CD by using the modulus function on embedding, minimizing pixel modification during the process, resulting in improved imperceptibility. In this final project, IHWT (Integer Haar Wavelet Transform) are used to keep imperceptibility high. To improve the security, cryptographic method NTRU is applied on the secret message before it is hidden in image. The result showed that the combination of NTRU, IHWT and modulus function yields good imperceptibility, with PSNR value above 40 dB while the stego image resist salt and pepper noise attack of 0,002% and contrast addition of maximum amount one]]>
<![CDATA[Uji Akurasi Algoritma Bipolar Slope One dan BW-Mine pada Sistem Rekomendasi ]]>
<![CDATA[Ali Akbar Lubis]]>;
<![CDATA[Ronsen Purba]]>;
<![CDATA[Megawaty Simamora]]>;
<![CDATA[Anna Agustiana]]>
<![CDATA[ Jurnal SIFO Mikroskil Vol 20, No 1 (2019): JSM Volume 20 Nomor 1 Tahun 2019 ]]>
Publisher : <![CDATA[Fakultas Informatika ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (606.83 KB)
|
DOI: 10.55601/jsm.v20i1.646
<![CDATA[The recommendation system is widely applied to various e-commerce. There are some problems that can cause the recommendation system to fail. This problem is about the massive vacuum of rating data (sparsity) and cold start. Therefore, the right recommendation method is needed to improve accuracy, so that the user can find the item according to desire.To achieve this goal, bipolar slope one is used to predict the rating. Bipolar slope one is used to predict the rating of an item. In predicting an item's rating, an item pattern is needed. This item pattern can be represented in the Assosiation Rule that found in the BW-Mine algorithm.The test was carried out with MAE involving 50 users of 200 items. The test results using MAE, obtained that sparsity has an influence on the accuracy of rating prediction generated in the recommendation system]]>
<![CDATA[Performance Enhancement and Accuracy of Artificial Neural Networks Using Particle Swarm Optimization for Breast Cancer Prediction ]]>
<![CDATA[Jimmy Nganta Ginting]]>;
<![CDATA[Ronsen Purba]]>;
<![CDATA[Erwin Setiawan Panjaitan]]>
<![CDATA[ InfoTekJar : Jurnal Nasional Informatika dan Teknologi Jaringan Vol 5, No 1 (2020): InfoTekJar September ]]>
Publisher : <![CDATA[Universitas Islam Sumatera Utara ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.30743/infotekjar.v5i1.2939
<![CDATA[Breast cancer is the one of leading causes of death among the women in many parts of the world. According to Global Cancer Observatory (GCO) data from WHO (2018) show that approximately 58,256 (16,7%) cancer cases were found in Indonesia out of a total of 348,809 cancer cases. The number of breast cancer patients throughout the world reached 42.1 per 100,000 population on average death rate of 17 per 100,000 inhabitants.Various ways have been used to find effective methods in the early detection of breast cancer. A prediction of breast cancer in early stage is very important in the medical world, which allows them to develop strategic programs that will help diagnose and reduce mortality rates from breast cancer. Performance enhancement and accuracy of artificial neural networks using particle swarm optimization is an effective solution for breast cancer prediction. The accuracy result was found 70% for training data and 96.1% for 30% prediction in this study. Previous studies only used the backpropagation algorithm to predict breast cancer and the result was 94.17%. Compared with previous study, there is an increase of 1.93% in combining Backpropagation with Particle Swarm Optimization.]]>
<![CDATA[COMBINATION OF ACO AND PSO TO MINIMIZE MAKESPAN IN ORDERED FLOWSHOP SCHEDULING PROBLEMS ]]>
<![CDATA[Sastra Wandi Nduru]]>;
<![CDATA[Ronsen Purba]]>;
<![CDATA[Andri]]>
<![CDATA[ INFOKUM Vol. 9 No. 2, June (2021): Data Mining, Image Processing and artificial intelligence ]]>
Publisher : <![CDATA[Sean Institute ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (1032.489 KB)
<![CDATA[The problem of scheduling flowshop production is one of the most versatile problems and is often encountered in many industries. Effective scheduling is important because it has a significant impact on reducing costs and increasing productivity. However, solving the ordered flowshop scheduling problem with the aim of minimizing makespan requires a difficult computation known as NP-hard. This research will contribute to the application of combination ACO and PSO to minimize makespan in the ordered flowshop scheduling problem. The performance of the proposed scheduling algorithm is evaluated by testing the data set of 600 ordered flowshop scheduling problems with various combinations of job and machine size combinations. The test results show that the ACO-PSO algorithm is able to provide a better scheduling solution for the scheduling group with small dimensions, namely 76 instances from a total of 600 inctances and is not good at obtaining makespan in the scheduling group with large dimensions. The ACO-PSO algorithm uses execution time which increases as the dimension size (multiple jobs and many machines) increases in a scheduled instance]]>
<![CDATA[SISTEM INFORMASI SIMPAN PINJAM PADA CU HARAPAN KITA BERBASIS CLIENT SERVER ]]>
<![CDATA[Adek Efnaldi Rambe]]>;
<![CDATA[Ronsen Purba]]>;
<![CDATA[Mendarissan Aritonang]]>
<![CDATA[ METHOMIKA: Jurnal Manajemen Informatika & Komputerisasi Akuntansi Vol. 1 No. 1 (2017): METHOMIKA: Jurnal Manajemen Informatika & Komputerisasi Akuntansi ]]>
Publisher : <![CDATA[Universitas Methodist Indonesia ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (481.915 KB)
|
DOI: 10.46880/jmika.Vol1No1.pp22-27
<![CDATA[Penerapan teknologi komputer dalam bidang simpan pinjam sangat banyak dijumpai, karena berguna untuk menyimpan, mengeluarkan, menghitung dan menampilkan keuangan pada unit simpan pinjam pada CU. Harapan Kita Belawan dalam kurun waktu tertentu. Pada CU. Harapan Kita Belawan semua pendataan administrasi perhitungan simpan pinjam dilakukan secara manual, yaitu dengan mencatat pada buku, baik data nasabah maupun pendapatan yang akan disetorkan kepada anggota koperasi. Hal ini menyebabkan banyak terjadi kesalahan data dan informasi. Dengan kondisi sistem yang masih manual, menyebabkan setiap bagian tidak dapat membagikan atau memberikan data kepada bagian yang lain. Bagian-bagian tersebut harus menginput kembali data yang akan digunakan. Kondisi tersebut mempengaruhi proses pelayanan khususnya pengolahan data, sehingga hal tersebut mengakibatkan kurang lancarnya mutu dan pelayanan yang diberikan kepada anggota koperasi.Dengan adanya komputer disuatu instansi maka segala yang ada di instansi tersebut dapat dengan mudah ditangani. Misalnya pada instansi atau bidang usaha yang bergerak dalam bidang layanan jasa simpan pinjam, dengan adanya komputer dan sistem pendataan, petugas akan mengetahui jumlah perhitungan simpanan dan pinjaman. Hal ini akan mempermudah petugas pengawasan untuk mengetahui pendapatan yang masuk ke perusahaan tersebut. Petugas juga mengetahui data-data pendapatan mulai dari data pinjaman nasabah sehingga dapat dicatat dan dihitung menggunakan komputer, dan dapat disimpan kedalam file yag diinginkan sehingga data akan mudah dicari apabila sewaktu-waktu dibutuhkan.]]>
<![CDATA[Analisis Perbandingan Algoritma ACO-TS dan ACO-SMARTER Dalam Menyelesaikan Traveling Salesman Problem ]]>
<![CDATA[Jimmy Peranginangin]]>;
<![CDATA[Ronsen Purba]]>;
<![CDATA[Arwin Halim]]>
<![CDATA[ JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 5, No 4 (2021): Oktober 2021 ]]>
Publisher : <![CDATA[STMIK Budi Darma ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.30865/mib.v5i4.3283
<![CDATA[The research conducted is the Comparative Analysis of the ACO-TS and ACO-SMARTER Algorithms in Solving the Traveling Salesman Problem where the problem to be solved is the traveling salesman problem (TSP). The purpose of this study is to hopefully be able to provide a comparison result of running time and the shortest distance between the ACO-TS algorithm and the ACO-SMARTER algorithm in solving the TSP. The test results show that the combination of the Ant Colony Optimization (ACO) algorithm and the Tabu Search (TS) algorithm is better in terms of achieving the optimum path and running time than the ACO and ACO-SMARTER algorithms in solving the Traveling Salesman Problem. The Tabu Search algorithm in the ACO algorithm acts as a controller for the routes that have been selected so that they are not processed again by the same ant. This will certainly make the ACO-TS algorithm faster in processing data because there is no data on the same route in the next round, where from 200 datasets the running time is obtained at ACO 11.5 seconds and the optimum distance is 76687, ACO SMARTER 8.5 seconds and the optimum distance is 74496 while the ACO-TS only takes 2.9 seconds and the optimum distance is 70558]]>
<![CDATA[Consumer Opinion Extraction Using Text Mining for Product Recommendations On E-Commerce ]]>
<![CDATA[Erlina Halim]]>;
<![CDATA[Ronsen Purba]]>;
<![CDATA[Andri Andri]]>
<![CDATA[ Indonesian Journal of Artificial Intelligence and Data Mining Vol 4, No 1 (2021): March 2021 ]]>
Publisher : <![CDATA[Universitas Islam Negeri Sultan Syarif Kasim Riau ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.24014/ijaidm.v4i1.10834
<![CDATA[This study aims to evaluate consumer opinions in text form on e-commerce to determine the accuracy of ratings given by consumers with opinions using text mining with the lexicon approach. The research data was obtained online using a crawling technique using the API provided by Shopee. The conditions of diverse opinions and use of non-standard words are challenges in processing opinions. Opinion must be processed normalization and repairs using dictionary of words before going to extract using lexicon approach. Dictionary of words contain opinions with weights that are worth 1 to 5 for positive opinions and are worth -1 to -5 for negative opinions. For each opinion will be classified using the maximum ratio of the weight of positive opinion compared to the weight of negative opinion. The classification of opinion produced is positive, negative or neutral. Opinion classification is then compared with the rating classification to work out the extent of accuracy. The comparison produces an accuracy of 80.34% by completing an opinion dictionary.]]>
<![CDATA[ENKRIPSI CITRA WARNA MENGGUNAKAN RUBIK’S CUBE DAN THREE CHAOTIC LOGISTIC MAP ]]>
<![CDATA[Ronsen Purba]]>;
<![CDATA[Frans Agus Purba]]>;
<![CDATA[Sari Fatmawati]]>
<![CDATA[ METHODIKA: Jurnal Teknik Informatika dan Sistem Informasi Vol. 2 No. 1 (2016): Maret 2016 ]]>
Publisher : <![CDATA[Universitas Methodist Indonesia ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.46880/mtk.v2i1.18
<![CDATA[An image encryption is a technique to protect the image secrecy from illegal accessing. One of the encryption algorithm that are used is Rubik's Cube where this algorithm used to permute the pixel in the color image. However, the key used in the Rubik's Cube is not random so the security of image is still weak. To further improve the security of the image, then used the chaotic systems for generating random keys. Excess of chaos is the sensitivity to initial conditions, behave randomly and do not have a recurring period. One of the chaos function used is Three Chaotic Logistic Maps where necessary keys generated are used in encryption and decryption proceses. The test results show that the algorithm Rubik's Cube and Three Chaotic Logistic Maps on the encryption and decryption processes do not cause an error in the original image and the result image. Neighboring pixels have a low coefficient indicating that the quality of the results encryption is good and resistant toward to attack by adding noise. The nature of chaotic sensitivity cause cipher image will not be recovered if the keys used in decryption are slightly diffrent from the encryption keys. that the decrypted image does not revert to plain image so the image obtained more secure.]]>
<![CDATA[PENGAMANAN DATA MENGGUNAKAN ENKRIPSI AES-256, PARITY CHECKER DAN GENERATOR MODULO ]]>
<![CDATA[Ronsen Purba]]>;
<![CDATA[Irpan Adiputra Pardosi]]>
<![CDATA[ METHODIKA: Jurnal Teknik Informatika dan Sistem Informasi Vol. 2 No. 2 (2016): Nopember 2016 ]]>
Publisher : <![CDATA[Universitas Methodist Indonesia ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.46880/mtk.v2i2.45
<![CDATA[Least Significant Bit (LSB) merupakan metode yang umum digunakan dalam steganografi karena mudah dimplementasikan dan imperceptibility yang baik. Namun teknik penyisipan yang bersifat sekuensial akan memudahkan penyerang dalam mendapatkan informasi yang tersimpan. Untuk mengatasi hal tersebut diperlukan metode penyisipan yang bersifat acak. Salah satu bentuk pengacakan yang cukup baik adalah generator modulo yang tergantung pada sebuah bilangan prima. Aspek imperceptibility dapat ditingkatkan dengan menerapkan parity checker. Kemudian untuk meningkatkan keamanan dapat dilakukan enkripsi terhadap data sebelum penyisipa. Penggabungan enkripsi, generator modulo dan parity checker menghasilkan sistem pengamanan data yang kuat. Penelitian ini menerapkan (1) enkripsi AES untuk mengacak data sebelum disisipkan ke dalam media penampung (citra sampul), (2) generator modulo digunakan untuk mengacak lokasi bit data yang akan disisipkan, dan (3) parity checker sebagai teknik penyisipan untuk meningkatkan imperceptability / fidelity sistem steganografi. Hasil pengujian yang dilakukan menunjukkan bahwa parity checker mampu meningkatkan keamanan metode LSB dalam aspek imperceptibility / fidelity dengan menghitung PSNR. Sistem ini juga diuji kemampuan pengembalian data dengan menambahkan noise pada citra stego dengan nilai probabilitas yang cukup tinggi. Terakhir penggunaan AES akan meningkatkan keamanan data yang disisipkan karena kunci yang digunakan bersifat rahasia.]]>
<![CDATA[Optimization Performance of Fuzzy K-Nn with Modifield Particle Swarm Optimization in Credit Risk Classification ]]>
<![CDATA[Wita Clarisa Ginting]]>;
<![CDATA[Ronsen Purba]]>;
<![CDATA[Arwin Arwin]]>
<![CDATA[ Jurnal Mantik Vol. 4 No. 2 (2020): Augustus: Manajemen, Teknologi Informatika dan Komunikasi (Mantik) ]]>
Publisher : <![CDATA[Institute of Computer Science (IOCS) ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.35335/mantik.Vol4.2020.956.pp1417-1423
<![CDATA[Credit risk is a risk due to the failure or inability of the customer to return the amount of credit obtained from the company and its interest according to a predetermined or scheduled period of time. The main task of the credit risk classification method is to provide a separation between those who have the potential to fail and those who have not failed in terms of credit payments. The k-Nearest Neighbor (kNN) method as the most popular, simple and easily implemented machine learning method can be used to classify credit risk. However, its success depends on the number of neighbors or neighbors (k) applied and the relationship between each data with a class is rigid (crisp) where each data only has a relationship with one class exclusively, while the other classes have no relationship at all. This study proposes the incorporation of the principles of fuzzy logic into k-NN to minimize the stiffness that results in a new method known as Fuzzy k-Nearest Neighbor or Fk-NN. However, the fuzzy strength factor (m) and the number of neighbors (k) as the fundamental determinants of Fk-NN which have a direct impact on the accuracy generated by the model, the determination is often not easy and difficult to control, so the Modified method is proposed Particle Swarm Optimization (MPSO) to be able to help Fk-NN find the best m and k values non-manually. The results of the classification of credit risk data are 1000 data, with 900 composition of training data (90%) and 100 data (10%) of test data using Fk-NN with MPSO producing accuracy reaching 92.4%, with the best k value is 7 and the best m value is 9.]]>
