{"title":"Pemanfaatan Metode SMOTE dan PSO Untuk Mengoptimalkan Tingkat Akurasi Klasifikasi Kepuasan Pelanggan","authors":"Arif Muqtadir, Dini Silvi Purnia","doi":"10.31294/ijcit.v8i1.11657","DOIUrl":"https://doi.org/10.31294/ijcit.v8i1.11657","url":null,"abstract":"Penyedia layanan telekomunikasi Indihome Witel Jakarta Selatan mempunyai keunggulan dan kekurangan yang dapat memberikan kepuasan dan ketidakpuasan yang membuat pelanggan berpindah ke penyedia layanan internet lain. Pengetahuan tentang klasifikasi konsumen yang puas dan tidak puas sangat dibutuhkan dalam mengambil keputusan. Tujuan dari penelitian ini adalah untuk menganalisa algoritma klasifikasi data mining mana yang menghasilkan nilai akurasi yang paling tinggi untuk klasifikasi kepuasan pelangan Indihome Jakarta Selatan antara algoritma Naïve Bayes, Support Vector Machine, Decission Tree. Dalam mengukur atau memprediksi tingkat kepuasan pelanggan dapat menggunakan algoritma klasifikasi data mining yaitu Naïve Bayes, Support Vector Machine, Decission Tree dengan penambahan fitur Metode Synthetic Minority Over-sampling Technique (SMOTE) dan teknik Particle Swarm Optimization (PSO).SMOTE berguna untuk menyeimbangkan data dan PSO untuk mengoptimalkan hasil nilai pada algoritma klasifikasi. Penelitian ini menghasilkan penggunaan Algoritma SVM + SMOTE + PSO menghasilkan Acurracy 98.57%, precision tertinggi 98.51%, recall tertinggi 98.51% dan AUC tertinggi 0.995 yang termasuk dalam excellent classification. Penggunaan SMOTE dan PSO pada algoritma klasifikasi terbukti berpengaruh meningkatkan accuracy algoritma SVM namun tidak secara signifikan. The South Jakarta Indihome Witel telecommunication service provider has advantages and disadvantages that can provide satisfaction and dissatisfaction that make customers switch to other internet service providers. Knowledge of the classification of satisfied and dissatisfied consumers is needed in making decisions. The purpose of this study is to analyze which data mining classification algorithm produces the highest accuracy value for the classification of South Jakarta Indihome customer satisfaction between the Naïve Bayes algorithm, Support Vector Machine, and Decission Tree. In measuring or predicting the level of customer satisfaction, data mining classification algorithms can be used, namely Naïve Bayes, Support Vector Machine, Decission Tree with the addition of the Synthetic Minority Over-sampling Technique (SMOTE) method if the data is unbalanced and the Particle Swarm Optimization (PSO) technique. SMOTE is useful for balancing data and PSO to optimize the value results in the classification algorithm. This research resulted in the use of the SVM + SMOTE + PSO algorithm resulting in an accuracy of 98.57%, the highest precision of 98.51%, the highest recall of 98.51% and the highest AUC of 0.995 which is included in the excellent classification. The use of SMOTE and PSO in the classification algorithm has proven to have an effect on increasing the accuracy of the SVM algorithm but not significantly.","PeriodicalId":397940,"journal":{"name":"IJCIT (Indonesian Journal on Computer and Information Technology)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123563165","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Audit Sistem Informasi menggunakan Framework Cobit 5 pada LPPM Universitas Semarang","authors":"Titis Handayani, B. V. Christioko","doi":"10.31294/ijcit.v8i1.11843","DOIUrl":"https://doi.org/10.31294/ijcit.v8i1.11843","url":null,"abstract":"Lembaga Penelitian dan Pengabdian kepada Masyarakat Universitas Semarang sebagai organisasi yang telah menerapkan teknologi informasi berupa sistem untuk mengelola data pelaksanaan kegiatan penelitian dan pengabdian kepada masyarakat. Namun, pada pemanfaatan sistem tersebut belum dilakukan pengukuran tingkat efektifitas dan kematangan pemanfaatan TI. Penelitian ini dilakukan dengan tujuan untuk mengevaluasi tata kelola berupa tingkat kematangan pemanfaatan TI dalam pengelolaan data kegiatan penelitian dan pengabdian kepada masyarakat. Pengukuran dilakukan menggunakan metode kerangka Cobit 5, dimana Domain yang digunakan adalah MEA (Monitor, Evaluate and Assess). Hasil pengukuran dari tingkat kematangan pemanfaatan TI di Lembaga Penelitian dan Pengabdian kepada Masyarakat USM saat ini semua berada pada level 1. Level tersebut memperlihatkan bahwa pada kondisi organisasi dalam hal ini adalah LPPM telah mengetahui tentang kegiatan dan proses teknologi informasi di organisasinya, hanya belum menyediakan aturan yang baku dalam pelaksanaan penerapan teknologi informasi dan mengembangkan suatu produk baru. Research and Community Service Institute at Semarang privateUniversity as an organization that has implemented information technology in the form of a system for managing data on the implementation of research activities and community service. However, the utilization of the system has not yet measured the level of effectiveness and maturity of IT utilization. This research was conducted with the aim of evaluating governance in the form of the maturity level of IT utilization in data management of research activities and community service . Measurements were carried out using the Cobit 5 framework method, where the Domain used was MEA (Monitor, Evaluate and Assess). The measurement results of the maturity level of IT utilization at the Semarang privateUniversity Research and Community Service Institute are currently all at level 1. This level shows that the organizational condition in this case is that LPPM already know about the activities and processes of information technology in the organization, but have not provided standard rules in implementing the application of information technology operational activities and develop a new product.","PeriodicalId":397940,"journal":{"name":"IJCIT (Indonesian Journal on Computer and Information Technology)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121625264","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Erika Mutiara, L. Ramdhani, Rusda Wajhillah, Jamal Maulana Hudin, A. Gunawan
{"title":"Penerapan Metode Naïve Bayes Dengan PSO Untuk Pemilihan Peminatan Jurusan Pada SMK","authors":"Erika Mutiara, L. Ramdhani, Rusda Wajhillah, Jamal Maulana Hudin, A. Gunawan","doi":"10.31294/ijcit.v8i1.15732","DOIUrl":"https://doi.org/10.31294/ijcit.v8i1.15732","url":null,"abstract":"Siswa SMP yang berencana melanjutkan jenjang pendidikannya ke SMK harus memutuskan program studi apa yang akan dipilih untuk melanjutkan pendidikannya kelak. Hal ini merupakan sesuatu yang cukup sulit diputuskan oleh kebanyakan siswa terutama bagi mereka yang tidak memiliki banyak referensi. Kebanyakan keputusan siswa, dipengaruhi oleh pendapat orang tua, teman atau figur yang diidolakan. Penelitian ini bertujuan untuk membangun sistem pendukung keputusan dengan metode Naive Bayes sebagai algoritma terpilih berbasis Particle Swarm Optimization (PSO) dalam pemilihan jurusan yang sesuai dengan kemampuan akademik siswa. Pada penelitian ini dilakukan optimasi metode Naive Bayes dengan menggunakan metode Particle Swarm Optimization untuk seleksi atribut untuk meningkatkan akurasi prediksi yang diaplikasikan terhadap data nilai calon siswa SMK. Setelah dilakukan pengujian dengan dua model, yaitu metode naive bayes dan optimasi naive bayes menggunakan particle swarm optimization hasil yang diperoleh menunjukan bahwa dengan menggunakan Naive Bayes nilai accucary yang dihasilkan sebesar 73,09% sedangkan pengujian dengan menggunakan optimasi Naive Bayes dengan Particle Swarm Optimization didapatkan nilai accurcy sebesar 80,43%. Sehingga kedua metode tersebut memiliki perbedaan tingkat akurasi sebesar 7,34%. The junior high school students who plan to continue their education to the vocational high school must decide what study program they will choose to continue their education later. This is something that is quite difficult for most students to decide, especially for those who don't have many references. Most students' decisions, sometimes influenced by the opinions of parents, friends or idolized figures. This study aims to build a decision support system using the Naive Bayes method as the selected algorithm based on Particle Swarm Optimization (PSO) in selecting majors according to students' academic abilities. In this study, the optimization of the Naive Bayes method was carried out by using the Particle Swarm Optimization method for attribute selection to improve the accuracy of predictions applied to prospective Vocational School student value data. After testing with two models, namely the Naive Bayes method and the Naive Bayes optimization using particle swarm optimization results obtained showed that using Naive Bayes the accucary value generated was 73.09% while testing using Naive Bayes optimization with Particle Swarm Optimization obtained values accurcy is 80.43%. So that both methods have different accuracy rates of 7.34%.","PeriodicalId":397940,"journal":{"name":"IJCIT (Indonesian Journal on Computer and Information Technology)","volume":"212 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115777692","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"User Experience Pengguna Perpanjangan Sim Online Aplikasi Digital Korlantas Menggunakan UEQ","authors":"Safira Herdianingsih, Oka Dwi Cahya","doi":"10.31294/ijcit.v8i1.13247","DOIUrl":"https://doi.org/10.31294/ijcit.v8i1.13247","url":null,"abstract":"Berdasarkan hasil observasi pada Google Play Store masih terdapat keluhan terkait sistem dan layanan perpanjangan SIM online pada aplikasi Digital Korlantas. Belum diketahui aspek-aspek terkait pengalaman pengguna yang mempengaruhi rendahnya rating aplikasi Digital Korlantas untuk menentukan area aspek pengalaman pengguna yang memerlukan peningkatan atau perbaikan pada aplikasi Digital Korlantas. Tujuan penelitian ini adalah Mengukur tingkat pengalaman pengguna terhadap layanan yang telah diberikan oleh aplikasi Digital Korlantas. Metode penelitian yang dilakukan menggunakan Data Kuantitatif dengan menyebarkan kuesioner yang mencakup beberapa pertanyaan terkait dengan apa yang dibahas, pertanyaan tersebut diberi skor 1-7. Kemudian hasilnya akan dihitung menggunakan metode User Experience Questionnaire. Dari hasil penilaian yang ada dapat disimpulkan skala Kebaruan pada aplikasi Digital Korlantas POLRI, bahwa pengguna menganggap layanan yang tersedia pada aplikasi Digital Korlantas POLRI kurang inovatif. Dengan adanya penelitian yang sudah dilakukan membuat aplikasi tersebut mempunyai evaluasi terhadap aplikasi yang sudah ada, agar aplikasi tersebut dapat nyaman digunakan oleh pengguna. Results Based on observations on the Google Play Store, there are still complaints regarding the system and services for adding an online SIM in the Digital Korlantas application. It is not yet known aspects related to user experience that affect the low rating of the Korlantas Digital application to determine aspects of user experience areas that require improvement or improvement in the Korlantas Digital application. The purpose of this study is to measure the level of user experience with the services provided by the Digital Korlantas application. The research method is quantitative data by distributing questionnaires that include several related to what is discussed, the questions are given a score of 1-7. Then the results will be calculated using the User Experience Questionnaire method. From the results of the existing assessment, it can open the Novelty scale on the POLRI Digital Korlantas application, that users consider the services available on the POLRI Digital Korlantas application to be less innovative. With the research that has been done, the application has an evaluation of existing applications, so that the application can be used by users","PeriodicalId":397940,"journal":{"name":"IJCIT (Indonesian Journal on Computer and Information Technology)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129378516","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Y. Farlina, Desi Susilawati, Tya Septiani Nurfauzia Koeswara, Agung Wibowo
{"title":"Aplikasi Persediaan Barang Pada Toko Dadun Menggunakan Model Prototype","authors":"Y. Farlina, Desi Susilawati, Tya Septiani Nurfauzia Koeswara, Agung Wibowo","doi":"10.31294/ijcit.v8i1.13713","DOIUrl":"https://doi.org/10.31294/ijcit.v8i1.13713","url":null,"abstract":"Di era kemajuan teknologi saat ini, orang dapat bertukar informasi dengan cepat meskipun mereka berjauhan. Pengelolaan persediaan menentukan kelancaran kegiatan usaha serta efektivitas dan efisiensi suatu usaha. Dalam sebuah bisnis, persediaan merupakan hal yang paling penting. Perusahaan perlu mempersiapkan dan mengelola persediaan secara teratur untuk menjaga agar proses produksi dan penjualan berjalan dengan lancar untuk menjalankan rencana dan tujuan mereka. Toko Dadun Sukabumi dalam proses pengelolaan persediaan barang belum memanfaatkan teknologi informasi, sehingga masih terjadi kesalahan dalam proses pencatatan stok barang. Selain itu, dalam pembuatan laporan persediaan barang mengalami kesulitan sehingga memerlukan waktu yang lama. Peneltiian ini bertujuan untuk membangun sistem informasi manajemen persediaan yang merupakan kombinasi dari teknologi informasi dan prosedur yang digunakan untuk manajemen persediaan, memantau data inventaris menggunakan sistem komputerisasi yang baik. Penelitian ini menggunakan model prototype dengan tahapan pengumpulan kebutuhan, membuat prototype, evaluasi prototype, mengkodekan sistem, menguji sistem, evaluasi sistem, dan menggunakan sistem. Penelitian ini menghasilkan sistem persediaan barang secara secara online untuk Toko Dadun. Dengan adanya sistem ini, mempermudah dalam pencatatan stok barang dan pendataan stok barang dapat terkontrol dengan baik. In today's era of technological advancement, people can exchange information quickly even though they are far apart. Inventory management determines the smooth running of business activities as well as the effectiveness and efficiency of a business. In a business, inventory is the most important thing. Companies need to prepare and manage inventory regularly to keep the production and sales processes running smoothly to carry out their plans and goals. Dadun Sukabumi Store in the process of managing inventory has not utilized information technology, so there are still errors in the process of recording stock of goods. In addition, in making inventory reports have difficulty so it takes a long time. This study aims to build an inventory management information system which is a combination of information technology and procedures used for inventory management, monitoring inventory data using a good computerized system. This study using a software model, namely a prototype model with the stages of gathering needs, making prototypes, evaluating prototypes, coding the system, testing the system, evaluating the system, and using the system. This research produces an online inventory system for the Dadun Store. With this system, it is easier to record stock of goods and collect data on stock items can be controlled properly.","PeriodicalId":397940,"journal":{"name":"IJCIT (Indonesian Journal on Computer and Information Technology)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128556329","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Sistem Pendukung Keputusan Pemilihan PPSU Terbaik Pada Kelurahan Bantarwaru Menggunakan Metode SAW","authors":"Mohamad Fajar Noer Aziz, Wulan Dari","doi":"10.31294/ijcit.v8i1.14243","DOIUrl":"https://doi.org/10.31294/ijcit.v8i1.14243","url":null,"abstract":"Sistem pemilihan petugas Penanganan Prasarana dan Sarana Umum (PPSU) pada Kelurahan Bantarwaru masih bersifat subjektif, seperti halnya kedekatan dengan kepala seksi/atasan. Dalam proses pengambilan keputusan pemilihan PPSU terbaik yang dilakukan oleh Lurah Bantarwaru mengalami kesulitan saat proses rekap dan belum adanya perankingan dari hasil penilaian akhir, sehingga sulit menentukan hasil kinerja PPSU terbaik. Metode Simpel Additive Weighting (SAW) merupakan cara efektif dalam menyeleksi PPSU terbaik, dengan menerapkan SPK sehingga dapat memutuskan hasil yang tepat dalam menseleksi PPSU terbaik. Penerapan metode SAW dilakukan dengan pembobotan terhadap kriteria karyawan. Dengan menggunakan metode SAW, permasalahan penilaian yang ada diharapkan menjadi lebih tepat, karena didasarkan pada perhitungan nilai kriteria dan bobot yang sudah ditentukan, sehingga mendapatkan hasil lebih efektif dan efisien karena waktu yang dibutuhkan dalam perhitungan lebih singkat. Hasil dari penelitian ini bahwa anggota PPSU terbaik pilihan Kelurahan Bantarwaru bernama Zainudin, dengan mendapatkan nilai total tertinggi pada angka 1.000. The selection system for officers who handle Public Infrastructure and Facilities (PPSU) in Bantarwaru Village is still subjective, such as closeness to section heads/superiors. In the decision-making process for selecting the best PPSU conducted by the Bantarwaru Lurah, they experienced difficulties during the recap process and there was no ranking from the results of the final assessment, so it was difficult to determine the best PPSU performance results. The Simple Additive Weighting (SAW) method is an effective way of selecting the best PPSU, by applying the SPK so that you can decide the right results in selecting the best PPSU. The application of the SAW method is carried out by weighting the employee criteria. By using the SAW method, the expected assessment problems will be more precise, because they are based on the calculation of predetermined criteria and weight values, so as to get more effective and efficient results because the time needed in the calculation is shorter. The results of this study are that the PPSU member who is the best choice for Bantarwaru Village is named Zainudin, with the highest total score at 1,000.","PeriodicalId":397940,"journal":{"name":"IJCIT (Indonesian Journal on Computer and Information Technology)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116146410","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Lia Mazia, Siti Fauziah, Susan Tahrima, Ega Nur Awangga
{"title":"Rancang Bangun Aplikasi StartUp Valetku","authors":"Lia Mazia, Siti Fauziah, Susan Tahrima, Ega Nur Awangga","doi":"10.31294/ijcit.v8i1.13435","DOIUrl":"https://doi.org/10.31294/ijcit.v8i1.13435","url":null,"abstract":"Dalam masyarakat saat ini, peningkatan kendaraan menciptakan tantangan bagi sistem parkir. Pengemudi mengalami kesulitan saat memarkirkan kendaraannya di tempat parkir, terutama di daerah padat seperti bandara dan pusat perbelanjaan. Sebagian besar tempat parkir tetap beroperasi dengan sistem parkir manual, dimana pengemudi mengemudi ke slot, mendapatkan tiket dari mesin di gerbang masuk, mencari slot yang tersedia, membayar parkir di loket, dan memasukan tiket berbayar ke tempat parkir. Melihat hal tersebut peneliti mencoba membuat StartUpValetku dengan menggunakan Mobile Application Development Life Cycle (MADLC) dan menggambarkan pemrograman berorientasi objek dengan Unified Modeling Language (UML). Untuk memudahkan dalam memahami setiap tahapan-tahapan dalam startup valetku, penulis menggunakan Mobile Application Development Life Cycle (MADLC) agar menghasilkan sebuah rancangan yang mampu mengelola data tersebut menjadi informasi sesuai dengan kebutuhan dengan lebih efektif dan efisien. Hasil dari pembuatan startup valetku diharapkan dapat membantu dan mempermudah pengendara untuk memarkirkan kendaraan mereka. Aplikasi ini juga diharapkan dapat membantu pihak pengelola tempat parkir untuk menarik pengunjung ke tempat parkir mereka. In today's society, the increase in vehicles creates challenges for the parking system. Drivers have difficulty parking their vehicles in parking lots, especially in congested areas such as airports and shopping malls. Most parking lots continue to operate under a manual parking system, where the driver drives to the slot, gets a ticket from the machine at the entrance gate. Seeing this, the researcher tries to create a Valetku StartUp using the Mobile Application Development Life Cycle (MADLC) and describes object-oriented programming with the Unified Modeling Language (UML). To make it easier to understand each stage in my valetku startup, the author uses the Mobile Application Development Life Cycle (MADLC) to produce a design that is able to manage this data into information according to needs more effectively and efficiently.The results of creating my valet startup are expected to help and make it easier for motorists to park their vehicles. This application is also expected to help the manager of the parking lot to attract visitors to their parking lot.","PeriodicalId":397940,"journal":{"name":"IJCIT (Indonesian Journal on Computer and Information Technology)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132840037","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Alif Muhamad Nisar, Erfian Junianto, Agung Baitul Hikmah
{"title":"Prediksi Penutupan Harga Saham PT Bank Rakyat Indonesia Dengan Metode Single Exponential Smoothing","authors":"Alif Muhamad Nisar, Erfian Junianto, Agung Baitul Hikmah","doi":"10.31294/ijcit.v8i1.14500","DOIUrl":"https://doi.org/10.31294/ijcit.v8i1.14500","url":null,"abstract":"Banyak investor yang masih ragu dalam melakukan investasi saham BRI yang tidak sesuai dengan harapan. Tujuan dalam penelitian ini adalah memprediksi harga penutupan saham BRI serta menerapkan metode Single Exponential Smoothing, dalam membantu memprediksi harga saham BRI. Periode observasi pada penelitian ini pada tanggal 1 januari 2022 hingga 30 April 2022. Data yang digunakan dalam penelitian ini menggunakan data sekunder yang diambil dari ir-bri. Untuk perhitungan yang dipakai dari rentang alpha 0.1 sampai 0.9 untuk mencari tingkat akurasi yang tinggi. Pada penelitian ini juga digunakan sampel dan metode peramalan Single Exponential Smoothing (SES) dan dalam proses menghitung akurasi peramalan ini digunakan metode MAPE (Mean Absolute Percentage Error). Hasil dari penelitian menggunakan metode Pemulusan Eksponensial Tunggal dan MAPE dengan nilai alpha 0.1, maka didapatkan hasil 4148 untuk SES dan selanjutnya untuk MAPE didapatkan hasil 0.7%. Sehingga penerapan metode SES cukup berguna dalam melakukan prediksi penutupan harga saham pada PT Bank Rakyat Indonesia, Tbk. Many investors are still unsure about investing in BRI shares that are not as expected. The purpose of this study is to predict the closing price of BRI shares and apply the Single Exponential Smoothing method to help predict BRI stock prices. The observation period in this study was from January 1 2022 to April 30 2022. The data used in this study used secondary data taken from ir-bri. For calculations used from the alpha range 0.1 to 0.9 to find a high level of accuracy. This study also used samples and the Single Exponential Smoothing (SES) forecasting method and in the process of calculating this forecast measurement the MAPE (Mean Absolute Percentage Error) method was used. The results of the study used the Single Exponential Smoothing method and MAPE with an alpha value of 0.1, the result was 4148 for SES and then for MAPE the result was 0.7%. So that the application of the SES method is quite useful in predicting the closing price of shares at PT Bank Rakyat Indonesia, Tbk.","PeriodicalId":397940,"journal":{"name":"IJCIT (Indonesian Journal on Computer and Information Technology)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125462222","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Analisis Prediksi Neural Network dan Support Vector Machine pada Multivariate Time Series Dataset","authors":"Mursalim Mursalim","doi":"10.31294/ijcit.v7i2.14443","DOIUrl":"https://doi.org/10.31294/ijcit.v7i2.14443","url":null,"abstract":"Data time series merupakan data yang sering diproduksi oleh manusia setiap harinya dari berbagai bidang baik bidang kesehatan, geoscience, biologi, astronomi dan lainnya. Namun, data time series ini banyak macamnya salah satunya adalah multivariate data dimana jenis data tersebut banyak digunakan oleh peneliti dalam berbagai penelitian. Metode Machine Learning yang digunakan adalah Neural Network atau NN dan Support Vector Machine atau SVM yang memiliki kelebihan masing-masing. CRISP-DM metodology akan menerapkan 5 tahapan mulai dari business understanding, data understanding, data preparation, modeling, evaluation dimana pada tahap modeling akan menggunakan 2 model yaitu Neural Network dan Support Vector Machine. hasil penelitian menunjukkan bahwa NN lebih dominan unggul dalam komparasi terbukti pada dataset polution on Beijing menghasilkan RMSE sebesar 0,13. Namun untuk data yang kecil seperti pada indeks saham Google NN mengalami large margin yaitu 2,98 diakrenakan data dan attribut yang terlalu sedikit untuk indeks saham","PeriodicalId":397940,"journal":{"name":"IJCIT (Indonesian Journal on Computer and Information Technology)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131107289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ayu Ainun A'ziziyyah, Bernadus Anggo Seno Aji, Muhammad Adib Kamali
{"title":"Rekomendasi Pemilihan Program Studi Menggunakan Support Vector Regression","authors":"Ayu Ainun A'ziziyyah, Bernadus Anggo Seno Aji, Muhammad Adib Kamali","doi":"10.31294/ijcit.v7i2.14120","DOIUrl":"https://doi.org/10.31294/ijcit.v7i2.14120","url":null,"abstract":"Salah jurusan saat kuliah berdampak pada mahasiswa akan malas dan mendapat nilai yang kurang memuaskan. Jurusan yang kurang seimbang dengan kemampuan mahasiswa mengakibatkan mahasiswa kurang mengerti materi atau bahkan tidak menyukai materi perkuliahan yang diberikan. Maka sangat penting bagi seorang siswa untuk memilih jurusan yang sesuai dengan bidang minat, bakat dan kemampuannya. Penelitian ini bertujuan untuk memberikan rekomendasi program studi bagi calon mahasiswa menggunakan metode Support Vector Regression dengan skenario penelitian berdasarkan input data yaitu menggunakan data nilai raport per semester dan data mean per mata pelajaran, skenario penelitian berdasarkan kernel yaitu menggunakan kernel RBF, Polynomial, dan Linear. Hasil akurasi terbaik didapatkan ketika menggunakan data nilai per semester dan kernel RBF, yaitu mendapat akurasi MAPE sebesar 5% dan MAE sebesar 0,16. Dan pada uji coba 10 sampel IPK tertinggi dari seluruh program studi didapatkan bahwa hasil dari rekomendasi 80% cocok dengan data asli. Wrong majors during college have an impact on students to be lazy and get unsatisfactory grades. Majors that are less balanced with student abilities result in students not understanding the material or even not liking the lecture material given. So it is very important for a student to choose a major that is in accordance with the areas of interest, talents and abilities. This study aims to provide study program recommendations for prospective students using the Support Vector Regression method with research scenarios based on input data, namely using report card scores per semester and mean data per subject, kernel based research scenarios using RBF, Polynomial and Linear kernels. The best accuracy results were obtained when using the value data per semester and the RBF kernel, which obtained an accuracy of 5% MAPE and 0.16 of MAE. And in the trial of the 10 highest GPA samples from all study programs, it was found that the results of the recommendations matched the original data 80%.","PeriodicalId":397940,"journal":{"name":"IJCIT (Indonesian Journal on Computer and Information Technology)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128359661","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}