Jurnal Transformatika
Not a member yet
319 research outputs found
Sort by
Penataan Ruang Kawasan Agropolitan di Kabupaten Semarang dengan Metode Artificial Neural Network
Full text linkKecamatan Bandungan dan Sumonowo merupakan wilayah di Kabupaten Semarang yang ditetapkan sebagai kawasan Agropolitan. Dalam beberapa survei dan penelitian di kawasan yang memiliki kesamaan administratif dan geografis tersebut menunjukan tren perubahan produktifitas pertanian akibat pembangunan. Pendekatan sistem informasi geografis digunakan dalam penelitian ini dengan alasan untuk bisa menganalisis objek spasial secara menyeluruh pada kawasan terkait. Penelitian ini berusaha memetakan penataan ruang secara objektif dengan metode Artificial Neural Network. Dimana hubungan antar objek spasial dikalkulasikan potensi perubahannya pada dua data spasial yang berbeda tempo. Hasilnya didapatkan bahwa simulasi objektif dengan metode ANN terhadap data spasial hasil klasifikasi menggunakan metode minimum distance diperoleh min validation error 0,0656. Hasil validasinya juga cukup baik, yaitu memperoleh prosentasi kebenaran 85,3% dan index kappa 0.80. Peta simulasi dihasilkan sampai pada tempo 2021. Dari peta simulasi didapatkan pengetahuan bahwa sistem pertanian terbuka akan terus mengalami pertumbuhan luasan secara positif dengan kisaran 0,015%. Adapun sistem pertanian tertutup akan terus mengalami penyusutan luasan pada kisaran 0,01%. Pengetahuan ini bisa menjadi alternaif solusi dalam mempertimbangkan implementasi rencana tata ruang dan wilayah pada kawasan terkait
ANALISA PREDIKSI MAHASISWA DROP OUT MENGGUNAKAN METODE DECISION TREE DENGAN ALGORITMA ID3 dan C4.5
Full text linkThe case of drop out at the Weleri STEKOM is often done by the campus. Drop out is a problem that is often done by Weleri STEKOM students because of a GPA of less than 2, Number of organizations followed, Tuition not paid for students and student who have exceeded the limit of 14 semesters. This study discusses predicting drop out students with C4.5 and ID3 decision tree methods that are useful to assist the campus in anticipating student dropouts. This study uses student data as many as 1087 students. Student data is divided into training data and testing data in order to obtain a model or rule in predicting DO students. Variable of this reseach contain V1(GPA) and then V2 (Distance beetween home and campus), V3(how long the lecture has been done), V4(Having a Job), V5(Family) and V6(school fee). This research The results of this study obtained 18 rules or rules for ID3 algorithm and 8 rules for C4.5 algorithm. The algorithm ID3 test results obtained an average of 95.17%, precision of 94.7% and recall of 96.18%, while for Decision Tree C 4.5 obtained an average of 96.45%, precision of 96.90% and recall of 95.38. This research prove that Decision using C4.5 is better for prediction of drop out students at STEKOM
Voice Over Internet Protocol Performance Evaluation in 6to4 Tunneling Network
Full text linkRegistry reported that their regional already in exhausted state. The IPv6 was proposed to substitute IPv4 network, but the implementation of this version cased many problems such as hardware compatibility. As temporary solution to this problem, 6to4 tunneling transition mechanism is introduced as one of many solutions. This mechanism used IPv4 network as communication media between two IPv6 networks. Thus, this kind of mechanism will affect the performance of Voice over Internet Protocol. VoIP demanded real-time communication by using UDP protocol between nodes. Unlike normal communication mode, real-time mode required data to be sent immediately ignoring the quality of data. This research evaluated the performance of 6to4 tunneling mechanism for Voice over Internet Protocol s communication between two nodes in native IPv6 networks.
Pemanfaatan Teknologi Machine Learning Untuk Klasifikasi Wilayah Risiko Kekeringan di Daerah Istimewa Yogyakarta Menggunakan Citra Landsat 8 Operational Land Imager (OLI)
Full text linkDrought is a natural disaster that occurs slowly and lasts a long time. Bantul and Gunung Kidul Regencies, Special Region of Yogyakarta are also areas affected by high drought risk. This happened because the area was the result of the construction of a cement factory and limestone mining along the Sewu Mountains. Prediction and classification of areas affected by drought can be done more accurately over large areas by extracting vegetation indices through remote sensing imagery. This research was conducted to provide information about the potential risk of drought in the region using Landsat 8 OLI spectral vegetation index data. Prediction or classification of drought potential using Artificial Neural Network. Vegetation index used in this study is NDVI, TCI, VCI, and VHI. Correlation results between vegetation indices showed the highest correlation occurred between the vegetation index TCI and VHI with the potential for a medium drought of 0.501 and the potential for a high drought of 0.684. Also obtained are the results of the classification of 9 villages that fall into the category of high drought potential (High Risk). Accuracy results and Kappa values indicate that Random Forest is the best method used with a breakdown of values of 99.91% and 99.81%, respectively. Spatial prediction results are performed using Inverse Distance Weighted (IDW) on vegetation index and prediction. Testing of spatial relationships between villages that have the potential for drought is done using Moran s I. analysis.
Location Based Service for improving Chabot Disaster Management Evacuator Palu
Full text linkThe devastating earthquake that struck Palu on the island of Sulawesi last September ripped through the Earth\u27s crust at a rare high speed, scientists have found. When the disaster is over, many natural disaster victims need immediate help. The call center provided is usually busy with services and complaints from victims of natural disasters. The greater the impact of natural disasters, the more information services that must be carried out. By using CEPAT chatbot for disaster evacuation in Palu, information about evacuation place can be given to victim by access it. Then when the victim shares their location, CEPAT will give the nearest evacuation place information using LBS improvement Chabot
Image Brightness Improvement Analysis Using HE, AHE, and ESIHE Comparison Methods
Full text linkImage improvement is the process to improve visual quality, from the original image to get optimal image results. The first category, this technique operates on the transformation of frequency selection and the second technique operates directly at the pixel level of the image. in this study the Exposure Sub-Image Histogram Equalization (ESIHE) technique will be enhanced with a brightness level to get visual image results. Then ESIHE is compared with other additional techniques, such as Histogram Equalization (HE), Adaptive Histogram Equalization (AHE) from the side of the image. Furthermore, for the even distribution of histograms, we will use ESIHE entropy calculations to show an increase in the image results that are more optimal when compared with HE and AHE. The visual image quality of each technique shows the strength of the method and the superiority of the other methods for various types of images
Decision Support System for Tuition Fee Deduction using Simple Additive Weighting
No full textOne of tools to get attention from student candidates was offered by UKRIDA was scholarship. Scholarship offered by UKRIDA for student candidates were tuition fee deduction and were implemented selectively according to the rules. The decision support system was built using SAW method as a simple and efficient method. The tuition fee deductions comprised of three types namely: National Exam score, school report card score and courses score (mathematics and English only). The weights for each criteria and sub-criteria was determined by applied provision in university. The result of the system was the number of deduction for each student candidate based on the major
Analisis Kinerja Algoritma Support Vector Machine (SVM) Guna Pengambilan Keputusan Beli/Jual Pada Saham PT Elnusa Tbk. (ELSA)
Full text linkStock is one of investing method that can improve the economy. Remote trading is one of the most popular trading method. Remote trading requires prediction of stock transaction signals to make it easier for traders to make decisions. Technical analysis is made easy with various indicators in analyzing stock price chart movements, such as Bollinger Bands, Pivot Point, MACD, Stochastic, ADX, and CCI, and then combined with Support Vector Machine (SVM) algorithm to classify sell/buy/hold classes, so we can obtain a pattern that is useful for predicting stock transaction signal decisions. The study was using WEKA software by analyzing the combination of indicators with the SVM algorithm where the object is historical data stocks of PT Elnusa Tbk. (ELSA). The highest profit obtained from this study is 28,02% which is the best model of the results of the data that is trained using non-aggressive sub sectors data using exponent value 2
Pembelajaran Text Preprocessing berbasis Simulator Untuk Mata Kuliah Information Retrieval
Full text linkPreprocessing is an important task and step in Information Retrieval. Information Retrieval (IR) is used to decide which documents in a collection must be taken to meet the user\u27s information needs. The comparison is done in 4 steps in preprocessing, namely Case folding, Tokenizing, Filtering, Stemming.All stages of the preprocessing are done manually during the teaching-learning process. The lecturer explained by writing and explaining one by one the correct words at each stage and students also manually reviewed each stage. Simulator is needed to help solve existing problems. The simulator has been successfully created by displaying 6 menus. The simulator has succeeded in outputting each stage according to the manual theory. From the results of the testing questionnaire,> 90% of respondents thought that using this simulator made it easier to work on the four stages of text pre-processing so that the time needed also became efficient
Implementasi Algoritma Cosine Similarity pada sistem arsip dokumen di Universitas Islam Sultan Agung
Full text linkArchiving in University that have not been well organized will cause a problems, the documents need for structuring and archives properly in the systems for the good standard a universities. The most importance of ease in finding the required archives is an important reason why it is necessary to develop an archive search system that can facilitate and improve the process of searching the archived document. Apllying cosine similarity algorithm in Information Systems is a solution for University to organizing archived documents, results from this reserach is the systems can show the relavant document from database list with precision 88.8% and recall 76.1% from all the data in database.