Claim Missing Document
Check
Articles

Found 8 Documents
Search

Data Warehouse Design For Sales Transactions on CV. Sumber Tirta Anugerah Syaputra, Muhammad Dwiky; Nazir, Alwis; Gusti, Siska Kurnia; Sanjaya, Suwanto; Syafria, Fadhilah
Jurnal CoreIT: Jurnal Hasil Penelitian Ilmu Komputer dan Teknologi Informasi Vol 8, No 2 (2022): December 2022
Publisher : Fakultas Sains dan Teknologi, Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (644.133 KB) | DOI: 10.24014/coreit.v8i2.19800

Abstract

Many data warehouses are implemented in companies engaged in retail, CV. Sumber Tirta Anugerah is one of the paint product retail companies that has not implemented it yet. As time goes by, the sales transaction data is getting more and more difficult to process because it is still stored in Microsoft Excel. This is a serious problem in utilizing historical data to assist in making a decision. It is difficult to store sales data because the data is quite large and a lot. Based on the above problems, a data warehouse design is needed for sales transaction data. This data warehouse design uses Kimball's nine-steps method and star schema. To perform the ETL process (extract, transform, and load) using Pentaho software. In this data warehouse design, Tableau software is used to visualize the processed data into a graph and dashboard report. The result of this research is a data warehouse design using nine steps and a star schema which gets a transformation response time of 4048 MS. 
Implementasi Algoritma Convolutional Neural Network (Resnet-50) untuk Klasifikasi Kanker Kulit Benign dan Malignant: Implementation of Convolutional Neural Network Algorithm (ResNet-50) for Benign and Malignant Skin Cancer Classification Gusti, Gogor Putra Hafi Puja; Haerani, Elin; Syafria, Fadhillah; Yanto, Febi; Gusti, Siska Kurnia
MALCOM: Indonesian Journal of Machine Learning and Computer Science Vol. 4 No. 3 (2024): MALCOM July 2024
Publisher : Institut Riset dan Publikasi Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.57152/malcom.v4i3.1398

Abstract

Kulit sebagai organ terluar yang menutupi seluruh bagian tubuh manusia rentan terhadap berbagi penyakit, salah satunya kanker kulit. Penggunaan teknologi malignant, khususnya Convolutional Neural Network (CNN) diangkat menjadi topik penelitian karena kemampuan CNN untuk secara otomatis mengenali fitur penting dalam klasifikasi citra medis kanker kulit. Oleh karena itu dilakukan penelitian pengklasifikasian penyakit kanker kulit benign (jinak) dan malignant (ganas) menggunakan algoritma CNN arsitektur ResNet-50 dengan dataset berupa 5000 data latih kanker kulit benign dan 4600 data latih kanker kulit malignant.Model CNN yang telah dirancang dengan epoch 50 menggunakan optimizer Adam dan batch size sebesar 54 serta melibatkan beberapa teknik augmentasi data guna meningkatkan keragaman dataset untuk kemudian model hasil perancangan diimplementasikan ke dalam tampilan sebuah website dengan menggunakan Flask sebagai kerangka kerja yang menghubungkan antara model deep learning dan website agar bisa diakses oleh pengguna. Metode pengujian blackbox dilakukan demi memastikan sistem dapat melakukan klasifikasi kanker kulit melalui input berupa citra medis kedalam 2 kelas yaitu benign dan malignant dengan baik serta didapatkan hasil akurasi model sebesar 94,88 % dan loss sebesar 13,24%.
Penerapan Metode Backpropagation Neural Network untuk Klasifikasi Penyakit Stroke Azhima, Mohd; Afrianty, Iis; Budianita, Elvia; Gusti, Siska Kurnia
KLIK: Kajian Ilmiah Informatika dan Komputer Vol. 4 No. 6 (2024): Juni 2024
Publisher : STMIK Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/klik.v4i6.1956

Abstract

Stroke is a non-communicable disease that can occur suddenly due to local or global disruption of brain function. The early symptoms of stroke are often difficult to recognize, causing many sufferers not to realize or feel the signs, so the death rate is quite high. This research aims to determine the ability of the Backpropagation Neural Network (BPNN) method in classifying stroke. The dataset used consists of 4891 medical records with stroke and non-stroke classes which include ten relevant variables (gender, age, hypertension, history of heart disease, BMI, blood sugar levels, and so on). This research runs three scenarios with the BPNN architecture model [19:25:1], [19:29:1], and [19:35:1] using a certain combination of variables, namely the comparison of training and testing data (90:10, 80 :20, 70:30), and learning rate 0.1; 0.01; 0.001. Test results with the highest average accuracy level of 96.14% were achieved with an architectural model of [19:29:1], a learning rate of 0.001, and a training and testing data distribution of 80:20. Based on testing, it can be concluded that BPNN is considered capable of classifying stroke
THE INFLUENCE OF MAJOR EXPERTISE COURSES ON ALUMNI EMPLOYMENT USING THE APRIORI METHOD Irsyad (Scopus ID: 57204261647), Muhammad; Iskandar, Iwan; Gusti, Siska Kurnia
Jurnal CoreIT: Jurnal Hasil Penelitian Ilmu Komputer dan Teknologi Informasi Vol 10, No 2 (2024): December 2024
Publisher : Fakultas Sains dan Teknologi, Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24014/coreit.v10i2.34144

Abstract

The role of alumni in university progress and quality is vital. This study used data from the tracer study application to analyze the relationship between skill courses and alumni employment. The data mining technique of association was employed to find linkages between different parameters. The Apriori algorithm was used to identify patterns that described the relationship between skill courses and alumni employment. The findings revealed that the most sought-after professions by alumni of the Informatics Engineering Study Program were educators, such as teachers and lecturers, with a support value of 18.7692%. Programmers were also in high demand, with a support value of 15.3846%. The subjects that were found to have the greatest influence on employment were Database, Computer Network, Computer Human Interaction, and Software Engineering. These findings provide valuable insights for the Informatics Engineering Study Program to prioritize and enhance these influential courses in terms of curriculum, teaching methods, and teaching materials, with the aim of improving the relevancy and quality of the courses in supporting alumni employment.
Implementation of XGBoost Ensemble and Support Vector Machine For Gender Classification of Skull Bones Ramadhani, Astrid; Afrianty, Iis; Budianita, Elvia; Gusti, Siska Kurnia
Bulletin of Informatics and Data Science Vol 4, No 1 (2025): May 2025
Publisher : PDSI

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61944/bids.v4i1.115

Abstract

Sex identification based on skull bones is an important step in forensic anthropology, especially in cases where unidentified human skeletons are found. Conventional methods such as DNA analysis are often used, but have limitations, especially when the bones are damaged, charred or decayed, making the analysis process difficult. This research applies XGBoost ensemble and Support Vector Machine for sex classification on skull bones. The purpose of this research is to handle complex data with many features and unbalanced data using the XGBoost ensemble method and Support Vector Machine (SVM). The data used consisted of 2,524 samples with 82 measurement features. Model performance was evaluated using accuracy, precision, recall, and F1 score metrics. The results showed that the combination of XGBoost and SVM methods, especially with the RBF kernel, was able to achieve accuracy of up to 91.52%. This finding proves that machine learning-based approaches can be an effective and reliable solution in supporting the forensic identification process
PENGARUH TEKNIK PENYEIMBANGAN DATA PADA KLASIFIKASI PENYAKIT NAFLD DENGAN ALGORITMA SVM Faska, Ridho Mahardika; Gusti, Siska Kurnia; Budianita, Elvia; Syafria, Fadhilah
Jurnal Informatika Teknologi dan Sains (Jinteks) Vol 7 No 2 (2025): EDISI 24
Publisher : Program Studi Informatika Universitas Teknologi Sumbawa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51401/jinteks.v7i2.5849

Abstract

Non-Alcoholic Fatty Liver Disease (NAFLD) merupakan penyakit hati kronis yang prevalensinya terus meningkat secara global, termasuk di Indonesia, dengan faktor risiko utama seperti obesitas, diabetes melitus, dan dislipidemia. Deteksi dini NAFLD menjadi tantangan penting karena metode konvensional seperti biopsi hati dan pencitraan memiliki keterbatasan dalam hal biaya, risiko invasif, dan kepraktisan. Penelitian ini bertujuan untuk mengembangkan model klasifikasi NAFLD menggunakan algoritma Support Vector Machine (SVM) dengan memanfaatkan dataset dari Kaggle yang terdiri dari 10 variabel dan 17.549 data. Untuk mengatasi masalah ketidakseimbangan kelas, diterapkan teknik oversampling seperti SMOTE, ADASYN, dan Random Oversampling (ROS) untuk melihat performa akurasi. Hasil penelitian menunjukkan bahwa SMOTE memberikan performa terbaik dengan akurasi tertinggi mencapai 78,70% pada kernel RBF, ROS dengan akurasi 78,18% dan ADASYN dengan akurasi 76,86%. Penelitian ini menyimpulkan bahwa pemilihan teknik oversampling data dan parameter yang tepat sangat penting dalam meningkatkan efektivitas model untuk menangani data tidak seimbang, sehingga dapat berkontribusi pada pengembangan metode deteksi NAFLD yang lebih efisien dan non-invasif.
Penerapan Information Gain Untuk Seleksi Fitur Pada Klasifikasi Jenis Kelamin Tulang Tengkorak Menggunakan Backpropagation Khair, Nada Tsawaabul; Afrianty, Iis; Syafria, Fadhilah; Budianita, Elvia; Gusti, Siska Kurnia
Bulletin of Computer Science Research Vol. 5 No. 4 (2025): June 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v5i4.637

Abstract

Forensic anthropology and skull analysis play a crucial role in the biological identification of individuals, including sex determination. This study aims to improve the accuracy of gender classification based on skull structure by combining the Information Gain feature selection method with the Backpropagation algorithm. The dataset used is the craniometric data compiled by William W. Howells, consisting of 2,524 samples with 85 measurement features. The preprocessing stage includes data selection, data cleaning, and normalization. Feature selection was conducted using the Information Gain method with three threshold values: 0.01, 0.05, and 0.1, resulting in 79, 46, and 38 selected features, respectively. The model was evaluated using the K-Fold Cross Validation method with K=10 and K=20. The highest accuracy of 93.91% was achieved at the 0.01 threshold using the Backpropagation architecture [79:119:1], a learning rate of 0.01, and K=20. These results demonstrate that feature selection using Information Gain enhances the performance of the Backpropagation model by eliminating irrelevant features and minimizing the risk of overfitting.
Clustering Keluarga Miskin Desa Bina Baru dengan Metode K-Medoids Amelia, Felina; Iskandar, Iwan; Gusti, Siska Kurnia; Haerani, Elin; Yusra, Yusra
Krea-TIF: Jurnal Teknik Informatika Vol 11 No 1 (2023)
Publisher : Fakultas Teknik dan Sains, Universitas Ibn Khaldun Bogor

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32832/krea-tif.v11i1.14104

Abstract

Kemiskinan di Indonesia terjadi di berbagai daerah, mulai pedesaan hingga perkotaan memiliki permasalahan kemiskinan masing – masing. Masalah kemiskinan juga dialami oleh Desa Bina Baru. Desa Bina Baru yang memiliki jumlah penduduk sebanyak 5.760 jiwa dengan total 1.742 keluarga, yang tersebar dalam 30 Rukun Tetangga (RT) dan 8 Rukun Warga (RW). Upaya dalam penurunan angka kemiskinan dapat dilakukan dengan berbagai cara, mulai pembangunan yang merata, penyaluran bantuan yang tepat sasaran, pemberian kebijakan yang tepat, dan lain sebagainya. Pengelompokan kemiskinan menjadikan salah satu upaya untuk menurunkan angka kemiskinan agar dapat memberikan informasi kepada pemerintahan daerah dalam memberikan kebijakan yang lebih tepat guna. Clustering merupakan teknik data mining yang bertujuan untuk mengelompokkan objek-objek data menjadi beberapa Cluster. Pada penelitian ini pengelompokkan dilakukan dengan teknik pengolahan data mining dengan algoritme K-Medoids dari data Desa Bina Baru tahun 2020 berjumlah 1.005. Hasil perbandingan perhitungan untuk Cluster 1 (kaya) sebanyak 527 penduduk, Cluster 2 (menengah) sebanyak 248 penduduk, dan Cluster 3 (miskin) sebanyak 225 penduduk, Hasil evaluasi dari algoritme k-Medoids adalah 0,991 yang menunjukan cluster yang dibentuk memberikan pengelompokan informasi yang baik. Hasil pengelompokan ini dapat dijadikan acuan untuk informasi kelompok keluarga miskin yang diperlukan pemerintah agar bantuan yang diberikan tepat sasaran.