Articles
Prediction of Rainfall and Water Discharge in The Jagir River Surabaya with Long-Short-Term Memory (LSTM)
Retzi Yosia Lewu;
Slamet Slamet;
Sri Wulandari;
Widdi Djatmiko;
Kusrini Kusrini;
Mulia Sulistiyono
Jurnal Riset Informatika Vol 5 No 3 (2023): Priode of June 2023
Publisher : Kresnamedia Publisher
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.34288/jri.v5i3.558
Flood disasters can occur at any time when the factors for the amount of river water discharge and rainfall intensity tend to be high, so preparations and ways of handling are needed to anticipate flood disasters quickly, precisely, and accurately for the Surabaya Public Works Service. One of the steps to predict and analyze the status of the flood disaster alert level is by calculating predictions based on rainfall and the amount of river water discharge. This study uses the Long-Short Term Memory (LSTM) algorithm to predict rainfall and river water discharge on the Jagir River in Surabaya. The LSTM method is a model commonly used for predictions based on time series data. The data obtained are rainfall data and water discharge on the Jagir River, Surabaya, which will be used as training and testing data to make predictions. The results of implementing the LSTM method using data training of 70% and data testing of 30% on rainfall data using the best epoch, namely at epoch ten by producing tests on data testing can have a Mean Absolute Error (MAE) performance of 4.5 and Root Mean Square Error (RMSE) of 9.7. Whereas the water discharge variable uses the best epoch, namely at epoch 75, by producing data testing data which can have a Mean Absolute Error (MAE) performance of 11.49 and a Root Mean Square Error (RMSE) of 9.63.
The Effect of Data Augmentation in Deep Learning with Drone Object Detection
Ariel Yonatan Alin;
Kusrini Kusrini;
Kumara Ari Yuana
IJCCS (Indonesian Journal of Computing and Cybernetics Systems) Vol 17, No 3 (2023): July
Publisher : IndoCEISS in colaboration with Universitas Gadjah Mada, Indonesia.
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.22146/ijccs.84785
 Drone object detection is one of the main applications of image processing technology and pattern recognition using deep learning. However, the limited drone image data that can be accessed for training detection algorithms is a challenge developing drone object detection technology. Therefore, many studies have been conducted to increase the amount of drone image data using data augmentation techniques. This study aims to evaluate the effect of data augmentation on deep learning accuracy in drone object detection using the YOLOv5 algorithm. The methods used in this research include collecting drone image data, augmenting data with rotate, crop, and cutout, training the YOLOv5 algorithm with and without data augmentation, as well as testing and analyzing training results.The results of the study show that data augmentation can't improve the accuracy of the YOLOv5 algorithm in drone object detection. Evidenced by the decreasing value of precision and mAP@0.5 and the relatively constant value of recall and F-1 score. This is caused by too much augmentation, which can cause a loss of important information in the data and cause noise or distortion.
Penggunaan Variabel Event dan Libur Sekolah Dalam Memprediksi Wisatawan Dengan Metode LSTM
Candra Rusmana;
Kusrini;
Kusnawi
JURNAL FASILKOM Vol 13 No 02 (2023): Jurnal FASILKOM (teknologi inFormASi dan ILmu KOMputer)
Publisher : Unversitas Muhammadiyah Riau
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.37859/jf.v13i02.4974
Event yang diadakan diberbagai daerah menjadi daya tarik tersendiri bagi wisatawan untuk datang ketempat tersebut. Liburan musiman sekolah juga menjadi agenda tahunan keluarga untuk pergi ke tempat wisata. Naik turunnya jumlah wisatawan yang datang ke Provinsi NTB memberikan dampak kepada pemerintah daerah, masyarakat sekitar tempat wisata dan pelaku usaha bidang pariwisata. Tujuan penelitian ini untuk melakukan pengujian terhadap variabel event tahunan dan libur sekolah. Datasetyang digunakan didapatkan dari website publik Provinsi NTB yaitu data.ntbprov.go.id dataset tersebut berupa histori jumlah kunjungan wisatawan setiap bulan, dari website disbudpar.ntbprov.go.id didapatkan dataset event tahunan dan dari website kalender pendidikan.com didapat dataset kalender akademik untuk liburan sekolah, dataset yang diambil dari setiap sumber diambil mulai dari tahun 2017 sampai tahun 2022. Dari semua dataset yang didapat bisa dimanfaatkan dalam menggali informasi untuk melakukan prediksi. Dalam melakukan prediksi digunakan Algoritma LSTM dengan menggunakan variabel histori wisatwan, event dan libu sekolah. Penggunaan variabel histori, event dan liburan menghasilkan kinerja MAPE sebesar 20.8% dengan penggunaan data training dan data testing 90/10. Hasil kinerja dengan variabel histori dan liburan saja menghasilkan kinerja MAPE sebesar 38,6%. sedangkan hasil dengan variabel histori dan event saja menghasilkan kinerja MAPE sebesar 23,81%. Ini menunjukan bahwa variabel event dan kalender liburan bisa dengan baik digunakan dalam melakukan prediksi terhadap kedatangan wisatawan di waktu berikutnya. Penelitian ini memperkenalkan pendekatan baru dalam memprediksi jumlah wisatawan dengan menggunakan variabel event tahunan dan kalender libur sekolah dengan menggunakan algoritma LSTM sebagai alat prediksi yang lebih canggih, yang sebelumnya belum banyak dieksplorasi dalam konteks prediksi pariwisata di Provinsi NTB.
PERBANDINGAN KINERJA METODE NAÏVE BAYES DAN SUPPORT MACHINE (SVM) DALAM ANALISIS KUALITAS BUTIR SOAL
Hidayatunnisa'i;
Kusrini;
Kusnawi
JURNAL FASILKOM Vol 13 No 02 (2023): Jurnal FASILKOM (teknologi inFormASi dan ILmu KOMputer)
Publisher : Unversitas Muhammadiyah Riau
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.37859/jf.v13i02.5087
Dalam melakukan analisis butir soal yang dilakukan proses pengumpulan, peringkas, dan penggunaan informasi dari jawaban siswa untuk membuat keputusan tentang setiap penilaian. Tujuan dari penilaian adalah untuk meningkatkan hasil belajar siswa serta memberikan informasi kepada siswa tentang kelebihan dan kekurangannya dalam mata pelajaran tertentu yang telah dipelajari. Oleh karena itu, dalam penelitian ini akan membahas tentang analisis perbandingan tingkat kinerja algoritma klasifikasi Naïve Bayes dan algoritma Support Vector Machine (SVM). Metode data mining untuk klasifikasi dapat digunakan untuk membantu meningkatkan kecepatan dan ketepatan dalam menganalisis butir soal sehingga akan didapatkan jenis soal yang diterima, direvisi, dan ditolak. Perbandingan kinerja algoritma Naive Bayes dan Support Vector Mahcine (SVM) bertujuan untuk mengukur tingkat akurasi dan lama waktu proses (execution time) dari masing-masing algoritma untuk mendapatkan algoritma terbaik yang akan diterapkan dalam membantu proses analisis butir soal. Data yang digunakan dalam penelitian sebanyak 50 dengan hasil jawaban siswa pada soal biologi dengan penggunaakn data training dan data testing 80:20. dengan menggunakan alat bantu bahasa pemrograman python, dapat disimpulkan bahwa algoritma Support Vector Machine (SVM) dengan K-fold Cross Validation sebesar 10 atau 10 kali tahap percobaan lebih unggul dibandingkan dengan Naïve Bayes.
Study Literature Study on Predicting Gold Prices using Machine Learning
Sri Wahyuningsih;
Kusrini;
Hanafi
DIELEKTRIKA Vol 10 No 2 (2023): DIELEKTRIKA
Publisher : Jurusan Teknik Elektro Fakultas Teknik Universitas Mataram
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.29303/dielektrika.v10i2.335
Gold is a metal that is believed to be able maintain prices and can be a medium of exchange that has different forms and functions. In addition, gold can be used as a long-term investment tool at various prices. Price variations can be influenced by many factors, so predicting gold prices can minimize risk. Gold price predictions are not only interesting for collectors but also interesting for analysts to discuss. Machine learning is one method that is often used to predict gold prices. In this study will discuss the study of literature based on the methods and results of previous literature. The purpose of this study is to determine the best performance of the methods that have been used and can be used as a reference in predicting gold prices. The purpose of writing this journal is to provide an overview as well the benefits of applying data mining techniques to machine learning. These benefits include development better understanding of machine learning, as well as improved decision making and technological innovation.
Penerapan Kombinasi Algoritma SVM-KNN dalam seleksi User SAKTI berdasarkan Hasil Kinerja Pegawai pada Kementerian XYZ
Syaiful Ramadhan;
Kusrini Kusrini;
Kusnawi Kusnawi
Jurnal Teknologi Informatika dan Komputer Vol 9, No 2 (2023): Jurnal Teknologi Informatika dan Komputer
Publisher : Universitas Mohammad Husni Thamrin
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.37012/jtik.v9i2.1716
Kementerian XYZ merupakan Kementerian dengan jumlah pegawai lebih dari 5.000 pegawai. Pada saat dibentuk tidak dilakukan pemetaan pegawai, hal ini mengakibatkan surplus jumlah pegawai, tidak terkecuali pada Biro Barang Milik Negara (BMN). Bagi sebuah organisasi, SDM yang berlimpah merupakan hal yang baik, namun perlu dilakukan penyeleksian pegawai agar dapat meningkatkan produktivitas sehingga keberhasilan organisasi dapat tercapai. Disamping itu, perbaikan sistem Administrasi Keuangan pemerintahan merupakan suatu keharusan yang diimbangi dengan pengembangan aplikasi terintegrasi Kementerian Keuangan yaitu Sistem Aplikasi Keuangan Tingkat Instansi (SAKTI). Dalam melakukan pengelolaan aset pada Biro BMN, setiap pegawai memiliki role user level kewenangan SAKTI dengan lingkup yang berbeda-beda. Penelitian ini bertujuan melakukan seleksi klasifikasi user berdasarkan hasil penilaian kinerja dengan penerapan metode Kombinasi algoritma SVM dan KNN menggunakan bahasa pemrograman Python. Berdasarkan pengujian dengan sampel data sebesar ±313 data pegawai dan 18 variabel pegawai dengan atribut target berupa kelayakan yaitu dipertahankan maupun dipertimbangkan, diperoleh hasil akurasi sebesar 94% pada Kernel SVM RBF; nilai K=5; metrik Euclidean; Dapat disimpulkan seleksi user aplikasi SAKTI menggunakan kombinasi algoritma SVM dan KNN dapat memberikan prediksi guna meningkatkan efektivitas dan efisiensi organisasi dalam penempatan pegawai yang sesuai dengan kompetensi pada Biro BMN Kementerian XYZ. Penelitian selanjutnya diharapkan dapat membandingkan kombinasi algoritma SVM dan KNN dengan metrik serta parameter yang lebih banyak.
Cross-site Scripting Attack Detection Using Machine Learning with Hybrid Features
Dimaz Arno Prasetio;
Kusrini Kusrini;
M. Rudyanto Arief
JURNAL INFOTEL Vol 13 No 1 (2021): February 2021
Publisher : LPPM INSTITUT TEKNOLOGI TELKOM PURWOKERTO
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.20895/infotel.v13i1.606
This study aims to measure the classification accuracy of XSS attacks by using a combination of two methods of determining feature characteristics, namely using linguistic computation and feature selection. XSS attacks have a certain pattern in their character arrangement, this can be studied by learners using n-gram modeling, but in certain cases XSS characteristics can contain a certain meta and synthetic this can be learned using feature selection modeling. From the results of this research, hybrid feature modeling gives good accuracy with an accuracy value of 99.87%, it is better than previous studies which the average is still below 99%, this study also tries to analyze the false positive rate considering that the false positive rate in attack detection is very influential for the convenience of the information security team, with the modeling proposed, the false positive rate is very small, namely 0.039%
Penerapan model InceptionV3 dalam klasifikasi penyakit ayam
Muhammad Salimy Ahsan;
Kusrini Kusrini;
Dhani Ariatmanto
JNANALOKA Vol. 04 No. 02 September Tahun 2023
Publisher : Lentera Dua Indonesia
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.36802/jnanaloka.2023.v4-no02-55-62
Chicken disease is one of the problems that can have a very significant impact on chicken farmers, in addition to having an impact on the farm itself, chicken disease can also have an impact on the surrounding environment. Lack of knowledge about the symptoms and diseases that occur in chickens, makes some chicken breeders treat and treat diseases in a traditional way. This method often takes a long time and is prone to errors. In this study, technology will be used to classify chicken diseases by utilizing a deep learning model from the Convolutional Neural Network (CNN) architecture, namely InceptionV3. In carrying out the process of classifying chicken diseases, using a dataset of chicken feces images with a number of 8067 Healthy, Salmonella, Coccidiosis, and Newcastle disease. In the research process, three experimental scenarios were carried out using 20 epochs, 50 epochs and 100 epochs. From the experimental results, using a value of 100 epochs produces the highest accuracy value with a value of 94.05%.
UNIVERSITY SERVICE WEBSITE QUALITY MEASUREMENT WITH WEBQUAL 4,0 (CASE STUDY: FACULTY OF BUSINESS AND ECONOMICS ISLAMIC UNIVERSITY OF INDONESIA)
Nur Hamid Sutanto;
Kusrini Kusrini;
Asro Nasiri
Jurnal Riset Informatika Vol. 3 No. 3 (2021): June 2021 Edition
Publisher : Kresnamedia Publisher
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.34288/jri.v3i3.82
The implementation of the website as a service medium has become very popular in the developments of universities, especially during the COVID-19 pandemic, which implements health protocols. With the existence of a website-based service information system, it is possible to continue running business processes in service to users without having to come and avoid direct or face-to-face interactions. FBE UII implements a Website Service Information System to serve students online, which replaces the manual system that is served directly and face-to-face. Firstly, it is implemented until now, there has never been a measurement of the quality of service that comes from its users. The measurement used the WebQual 4.0 method which measures the user's assessment of 3 areas, namely usability, information quality, service interaction quality which affects user satisfaction with the parameter of the users using the service website. This research was conducted at the FBE UII. Research data collection used a questionnaire. The population was FBE UII students who had used the service website. The results of this study were the conclusions of successful website implementation to replace the previously used system, the quality of website-based services on information systems was good, the perception of user ratings was easy to interact with, and recommendations for strategic steps, namely migrating other services by utilizing integrated service websites.
Prediction of Rainfall and Water Discharge in The Jagir River Surabaya with Long-Short-Term Memory (LSTM)
Retzi Yosia Lewu;
Slamet Slamet;
Sri Wulandari;
Widdi Djatmiko;
Kusrini Kusrini;
Mulia Sulistiyono
Jurnal Riset Informatika Vol. 5 No. 3 (2023): June 2023
Publisher : Kresnamedia Publisher
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.34288/jri.v5i3.239
AbstractFloods can occur at any time if the amount of river water discharge and rainfall intensity tends to be high, so preparations and ways of handling are needed to anticipate flooding quickly, precisely, and accurately for the Surabaya City Public Works Service. One of the steps to predict and analyze the status of the flood disaster alert level is to calculate predictions based on rainfall and the amount of river water discharge. This study uses the Long-Short Term Memory (LSTM) algorithm to predict using a time series dataset of rainfall and river water discharge in the Jagir River, Surabaya. This data is used to make predictions with the proportion of 70% training data and 30% testing data. Data normalization is performed in intervals of 0 and 1 using a min-max scaler and activated using ReLU (Rectified Linear Unit) and Adam Optimizer. The process continues by repeating the process to enter iterations, or epochs until it reaches the specified epoch (n). The data is then normalized to their original values and visualized. The model was evaluated and produced acceptable performance evaluation results for the rainfall variable, namely at epoch (n) = 75 for training data, namely a score of 0.054 for MAE and 0.099 for RMSE. In contrast, data testing was given a score of 0.041 for MAE and 0.091 for RMSE. As for the water discharge variable, the performance evaluation shows the difference between the training and testing data. Results of training data MAE = 11.10 and RMSE=18RMSE =18.61.61 at epoch (n) = 150. Results of data testing MAE = 11.37 and RMSE = 21.08 at epoch (n) = 100. These results indicate an anomaly that needs to be discussed in further research.