Articles
<![CDATA[APLIKASI METODE GOLDEN SECTION UNTUK OPTIMASI PARAMETER PADA METODE EXPONENTIAL SMOOTHING ]]>
<![CDATA[Mahkya, Dani Al]]>;
<![CDATA[Yasin, Hasbi]]>;
<![CDATA[Mukid, Moch. Abdul]]>
<![CDATA[ Jurnal Gaussian Vol 3, No 4 (2014): Jurnal Gaussian ]]>
Publisher : <![CDATA[Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (633.472 KB)
|
DOI: 10.14710/j.gauss.v3i4.8071
<![CDATA[Forecasting is predicting the activities values that have been previously known. One of the methods that can be used to predict is Exponential Smoothing. In this study, exponential smoothing method used is Single Exponential Smoothing (SES), Holt Double Exponential Smoothing (DES) and Triple Exponential Smoothing Holt-Winter (TES) Additive and Multiplicative models. Data used is value of Central Java Export from the period January 2006 until December 2013. There is some weighting parameters were evaluated in this method in order to produce a minimum error. Trial error method is used to obtain the weighting parameters. For SES method parameters evaluated were the parameters α, in DES method there are α and γ. And TES method there are α, γ and β. The value that will be minimize is Persentage Mean Absolute Error (MAPE). This study used the Golden Section method to find the parameter values that minimize the weighting function of MAPE. And built a Graphical User Interface (GUI) MATLAB in order to facilitate the analysis process. The Golden Section analysis found the best model is the TES Holt Winters Additive because it has a minimum value of MAPE. With Use the TES Holt Winters Additive will continue to predict the value of exports of Central Java 12 periods ahead with weighting parameters that minimize MAPE. Keywords : Exponential Smoothing, Graphical User Interface (GUI), Export, Golden Section, Predict]]>
<![CDATA[ANALISIS FAKTOR UNTUK MENGANALISIS VARIABEL PENDUDUK MISKIN ]]>
<![CDATA[Anggraini, Dian]]>;
<![CDATA[Al Mahkya, Dani]]>;
<![CDATA[Fitriawati, Andi]]>;
<![CDATA[Siahaan, Radot MH]]>
<![CDATA[ MAp (Mathematics and Applications) Journal Vol 2, No 1 (2020) ]]>
Publisher : <![CDATA[Universitas Islam Negeri Imam Bonjol Padang ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (520.631 KB)
|
DOI: 10.15548/map.v2i1.1636
<![CDATA[Data kemiskinan Indonesia umumnya menggunakan pengukuran basic needs approach. Berdasarkan data Survei Sosial Ekonomi Nasional (SUSENAS) yang dilakukan oleh Badan Pusat Statistik (BPS) setiap tahun, BPS membagi kriteria kesejahteraan masyarakat menjadi tiga kelompok berdasarkan kelompok pengeluaran, yaitu 40% kebawah, 40% tengah dan 20% keatas. Kelompok pengeluaran 40% kebawah bisa dikatakan sebagai penduduk miskin. Kelompok inilah yang akan diamati dan dilakukan analisis. Data BPS 10 tahun terakhir menyebutkan bahwa kemiskinan Provinsi Lampung terus mengalami penurunan setiap tahunnya dari tahun 2008 sebesar 22,19% sampai tahun 2018 menjadi 13,14%. Analisis faktor merupakan salah satu teknik untuk mengombinasikan variabel dengan tujuan menciptakan kelompok variabel baru. Analisis faktor sendiri dibagi menjadi dua, yaitu analisis faktor eksploratori dan konfirmatori. Informasi yang diperoleh dari data SUSENAS 2018 akan digunakan dalam analisis dan harapannya variabel baru yang didapatkan dengan analisis faktor eksploratori bisa menyederhanakan variabel yang selama ini digunakan oleh BPS. Berdasarkan hasil analisis faktor eksploratori yang telah dilakukan dapat disimpulkan bahwa faktor baru yaitu faktor kebutuhan tambahan dan faktor kebutuhan utama yang terbentuk dari 5 variabel awal bisa digunakan untuk menggambarkan kondisi kelompok masyarakat dengan kelompok pengeluaran 40% kebawah. Faktor kebutuhan tambahan terdiri dari status kepemilikan rumah sendiri, memakai BPJS dan bisa baca tulis. Sedangkan faktor kebutuhan utama terdiri dari sumber air bersih yang digunakan dan pengeluaran untuk makan.AbstractThe Poverty data Indonesia usually use the basic needs approach. Based on the national socio economic data (SUSENAS) conducted by the Central Bureau of Statistics (BPS) every year. BPS divide criteria community welfare into three groups based on the expenditure, namely 40% down, 40% middle and 20% above. Level Expenditure group 40 % down can be as poor group. These groups are the ones to be observed and analysis. BPS data for the last 10 years states that poverty in Lampung Province continues to decrease every year from 2008 by 22.19% until 2018 to 13.14%. Factor analysis is one technique for combining variables with the aim of creating a new group of variables. Factor analysis itself is divided exploratory and confirmatory factor analysis. Information obtained from the 2018 SUSENAS data will be used in the analysis and it is hoped that new variables obtained by exploratory factor analysis can simplify the variables that have been used by BPS. Based on the results of exploratory factor analysis that has been carried out, it can be concluded that the new factors, namely the additional needs factor and the main needs factor formed from the 5 initial variables, can be used to describe the condition of community groups with expenditure groups 40% down. The additional needs factor consists of the ownership status of the house itself, using BPJS and being able to read and write. While the main needs factor consists of the source of clean water used and expenditure for food.]]>
<![CDATA[PEMODELAN KEBERGANTUNGAN DALAM MENGKONSTRUKSI DISTRIBUSI BIVARIAT COPULA FRANK PADA DATA MARGINAL DISKRIT MELALUI TRANSFORMASI NORMAL STANDAR DAN JITTERS ]]>
<![CDATA[Andi - Fitriawati]]>;
<![CDATA[Dani Al Mahkya]]>;
<![CDATA[Radot MH Siahaan]]>;
<![CDATA[Dian Anggraini]]>
<![CDATA[ VARIANCE: Journal of Statistics and Its Applications Vol 2 No 1 (2020): VARIANCE: Journal of Statistics and Its Applications ]]>
Publisher : <![CDATA[Statistics Study Programme, Department of Mathematics, Faculty of Mathematics and Natural Sciences, University of Pattimura ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.30598/variancevol2iss1page1-13
<![CDATA[Data diskrit merupakan data empirik hasil realisasi variabel acak diskrit maupun kontinu. Ketika memiliki dua jenis data diskrit, seringkali ingin dikonstruksi distibusi bivariatnya untuk berbagai keperluan, baik fungsi peluang maupun fungsi distribusinya. Namun, saat data yang dimiliki terdapat kebergantungan, maka mengkonstruksi distibusi bivariatnya tidaklah mudah. Oleh sebab itu, digunakan Copula. Permasalahan lain timbul ketika data yang dimiliki tidak hanya memiliki kebergantungan tetapi juga berasal dari marginal diskrit. Berdasarkan teorema Sklar, penggunaan Copula dalam mengkonstruksi distribusi bivariat pada marginal diskrit akan menghasilkan suatu Copula C yang tidak unik. Akibatnya akan menimbulkan interprestasi yang tidak jelas, terutama untuk sifat kebergantungannya. Oleh sebab itu, diperlukan suatu teknik untuk mengkonstruksi distribusi bivariat dari data tersebut, yaitu dengan mengkontinukan distribusi marginalnya. Mengkontinukan distribusi marginalnya dilakukan melalui transformasi normal standar dan jitters. Hasil trasnformasi mampu mempresentasikan data aslinya. Hal ini terlihat dari perilaku penyebaran data dan ukuran kebergantungan dari data hasil transformasi dengan data aslinya adalah sama. Ukuran kebergantungan yang digunakan, yaitu Korelasi Pearson dan Kendall’s tau. Selanjutnya, hasil transformasi ini kemudian digunakan untuk mengkontrusksi distribusi bivariat dari data yang dimiliki menggunakan Copula. Copula yang digunakan adalah Copula Frank dengan asumsi bahwa data tidak memiliki kebergantungan ekor atas maupun bawah. Jadi, fungsi peluang bivariat dan/atau fungsi distribusi bivariat dari data hasil transformasi mempresentasikan fungsi peluang bivariat dan/atau fungsi distribusi bivariat dari data aslinya. Seluruh prosesnya akan diilustrasikan melalui data simulasi.]]>
<![CDATA[EXTRA TREES METHOD FOR STOCK PRICE FORECASTING WITH ROLLING ORIGIN ACCURACY EVALUATION ]]>
<![CDATA[Dani Al Mahkya]]>;
<![CDATA[Khairil Anwar Notodiputro]]>;
<![CDATA[Bagus Sartono]]>
<![CDATA[ MEDIA STATISTIKA Vol 15, No 1 (2022): Media Statistika ]]>
Publisher : <![CDATA[Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.14710/medstat.15.1.36-47
<![CDATA[Stock is an investment instrument that has risk in its management. One effort to minimize this risk is to model and make further forecasts of stock price movements. Time series data forecasting with autoregressive models is often found in several cases with the most popular approach being the ARIMA model. The tree-based method is one of the algorithms that can be used to forecast both in classification and regression. One ensemble approach to tree-based methods is Extra Trees. This study aims to forecast using the Extra Trees algorithm by evaluating forecasting accuracy with Rolling Forecast Origin on BRMS stock price data. Based on the results obtained, it is known that Extra Trees produces a fairly good accuracy for forecasting up to 6 days after training data with a MAPE of less than 0.1%.]]>
<![CDATA[Peramalan Cryptocurrency dengan Autoregressive Integrated Moving Average (ARIMA) dan Risiko Kerugian dengan Value at Risk (VaR) ]]>
<![CDATA[Amalia Listiani]]>;
<![CDATA[Dani Al Mahkya]]>
<![CDATA[ Journal of Science and Applicative Technology Vol 6 No 2 (2022): Journal of Science and Applicative Technology December Chapter ]]>
Publisher : <![CDATA[Lembaga Penelitian dan Pengabdian Masyarakat (LPPM), Institut Teknologi Sumatera, Lampung Selatan, Lampung, Indonesia ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.35472/jsat.v6i2.904
<![CDATA[Blockchain is a technology that is used for recording digital transactions that are interconnected and cannot be changed. Cryptocurrencies use blockchain technology, which has advantages due to a high level of security, low fees, and a high return on investment. One of the most popular cryptocurrencies and one that has a high market cap is Bitcoin. High volatility carries the risk of large losses. So it is necessary to analyze the risk of loss and forecast Bitcoin. Forecasting is carried out using the Autoregressive Integrated Moving Average (ARIMA) model, which is then carried out by risk analysis using Value at Risk (VaR) using the Historical Data method. Based on the research results, ARIMA [4,1,2] was great for predicting Bitcoin, with a Mean Absolute Percentage Error (MAPE) of 6%. Based on the results of research with Value at Risk (VaR), investors have a maximum loss tolerance of 5.86% and there is a 5% possibility that the losses will exceed 5.85%.]]>
<![CDATA[Pemodelan dan prediksi jumlah penumpang pelabuhan bakauheni selama periode tsunami Selat Sunda menggunakan autoregressive integrated moving average ]]>
<![CDATA[Dani Al Mahkya]]>;
<![CDATA[Dian Anggraini]]>;
<![CDATA[Andi Fitriawati]]>;
<![CDATA[Radot MH Siahaan]]>
<![CDATA[ Journal of Science and Applicative Technology Vol 4 No 1 (2020): Journal of Science and Applicative Technology June Chapter ]]>
Publisher : <![CDATA[Lembaga Penelitian dan Pengabdian Masyarakat (LPPM), Institut Teknologi Sumatera, Lampung Selatan, Lampung, Indonesia ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (536.671 KB)
|
DOI: 10.35472/jsat.v4i1.266
<![CDATA[Bakauheni Port is a ferry port located in Bakauheni District, South Lampung. This port is one of the major ports located on Sumatera island connecting Sumatera and Java and is located in the Sunda Strait. The tsunami that occurred in the Sunda Strait on December 22, 2018 indirectly affected the sea crossing node, especially the Bakauheni-Merak route. This can lead to changes in time series data patterns. The phenomenon is expected to be captured through a mathematical modeling that can be used as a decision making in the future. The purpose of this study was to model and predict the number of Bakauheni port passengers during the Sunda Strait Tsunami period using the Autoregressive Integrated Moving Average (ARIMA). The ARIMA approach uses past information as a basis for modeling. Based on visual information on the number of Bakauheni Port passengers, there was an increase in December in general. Other information is that there are seasonal patterns that occur with a span of 7 days. This was indicated by the pattern of repeated increases in the number of passengers every Sunday. After the tsunami, the number of passengers decreased for 2 days. In the 3 days after the Tsunami or during the Christmas holiday on December 25, 2018, the number of passengers has increased again. Based on the analysis and discussion that has been done, the best time series model obtained is ARIMA([5],1,2)(2,1,0)7 with a Mean Absolute Percentage Error (MAPE) of 9.55%.]]>
<![CDATA[Pemodelan Pergerakan Harga Saham Bakrie Menggunakan Metode ARIMA (Autoregressive Integrated Moving Average) ]]>
<![CDATA[Nabilah Syafitri]]>;
<![CDATA[Aziza Indah Putri]]>;
<![CDATA[Dinda Citra Utami]]>;
<![CDATA[Deva Dery]]>;
<![CDATA[Shandika Bayu Perkasa]]>;
<![CDATA[Dani Al Mahkya]]>
<![CDATA[ Indonesian Journal of Applied Mathematics Vol 1 No 1 (2020): Indonesian Journal of Applied Mathematics Vol. 1 No. 1 October Chapter ]]>
Publisher : <![CDATA[Lembaga Penelitian dan Pengabdian Masyarakat (LPPM), Institut Teknologi Sumatera, Lampung Selatan, Lampung, Indonesia ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
<![CDATA[Peramalan digunakan untuk memprediksi sesuatu yang akan terjadi di masa mendatang sehingga tindakan yang tepat dapat dilakukan. ARIMA merupakan salah satu metode peramalan runtun waktu yang dikembangkan dimana data pengamatan dalam sebuah data runtun waktu diasumsikan berhubungan satu sama lain secara statistik. Tujuan dari penelitian ini adalah untuk membuat model dan meramalkan harga saham PT. Bakrie Sumatera Plantations menggunakan metode ARIMA. Data yang digunakan adalah data harga saham PT. Bakrie Sumatera Plantations selama periode 2 Februari 2020 sampai 27 Juli 2020. Model akhir yang diperoleh adalah ARIMA(1,1,1)(1,0,1)3.]]>
<![CDATA[Pelatihan Penerapan Sistem Integrasi Data Kependudukan Sederhana (SIDaKS) Di Kecamatan Kota Agung Timur Tanggamus ]]>
<![CDATA[Dani Al Mahkya]]>;
<![CDATA[Fery Widhiatmoko]]>;
<![CDATA[Dian Anggraini]]>;
<![CDATA[Tirta Setiawan]]>;
<![CDATA[Febri Dwi Irawati]]>;
<![CDATA[Meida Cahyo Untoro]]>
<![CDATA[ Jurnal Pengabdian kepada Masyarakat Radisi Vol 2 No 1 (2022): April ]]>
Publisher : <![CDATA[Yayasan Kajian Riset dan Pengembangan RADISI ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.55266/pkmradisi.v2i1.68
<![CDATA[System-based data recording is indispensable in all fields, including government agencies. One of the important points in data recording is system integration. Data integration has the advantage of making the flow of organizational information better. One of the problems that arise in this Community Service activity partner is the unavailability of a population data integration system from each village/village to the sub-district. This activity aims to conduct training, mentoring and demonstration of the proposed integration system. The solution that will be offered is to conduct training and socialization related to population data management and integration. This training and socialization will use Ms. Excel and Google Sheet in the process. The method used in implementing community service activities is training and mentoring as well as demonstrations related to the Implementation of the Simple Population Data Integration System (SIDaKS) in Kota Agung Timur District, Tanggamus. The training was carried out in the East Kota Agung District hall, Tanggamus. Participants who attended the activity were representatives of each village/village in the Kota Agung Timur District. There are several steps taken to support the implementation of activities. The activity took place smoothly in accordance with the applicable health protocol. And it will be held on September 30, 2021 at the Kota Agung Timur , Tanggamus]]>
<![CDATA[Pelatihan Pembuatan Laman Pekon Informatif Sebagai Sarana Promosi Potensi Lokal Wilayah Kota Agung Timur ]]>
<![CDATA[Fauzi, Rifky]]>;
<![CDATA[Pribadi, Aswan Anggun]]>;
<![CDATA[Edriani, Tiara Shofi]]>;
<![CDATA[Mahkya, Dani Al]]>;
<![CDATA[Listriani, Amalia]]>;
<![CDATA[Mahrani, Dwi]]>
<![CDATA[ TeknoKreatif: Jurnal Pengabdian kepada Masyarakat Vol 3 No 1 (2023): TEKNOKREATIF : Jurnal Pengabdian kepada Masyarakat ]]>
Publisher : <![CDATA[Lembaga Penelitian dan Pengabdian Kepada Masyarakat (LP2M), Institut Teknologi Sumatera, Lampung, Indonesia. ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.35472/teknokreatif.v3i1.780
<![CDATA[Sejak tahun 2003 pemerintah melalui Instruksi Presiden Nomor 3 Tahun 2003 tentang e-Government mendorong seluruh pelayanan publik agar dapat diakses secara 24 jam dari manapun dan kapanpun. Hal ini mendorong satuan pemerintahan terkecil yakni desa untuk dapat beradaptasi, salah satunya adalah dengan menyediakan laman desa yang dapat diakses oleh masyarakat. Fungsi lain dari laman desa adalah sebagai sarana mempromosikan potensi-potensi desa. Meski aturan mengenai digitalisasi layanan pemerintah ini telah bergulir sejak lama, beberapa desa (pekon) khususnya pekon-pekon di wilayah Kecamatan Kota Agung Timur belum memiliki laman desa. Padahal desa-desa tersebut memiliki banyak potensi baik kepariwisataan maupun hasil perkebunan dan perikanan. Untuk mendukung hal tersebut, kegiatan pengabdian berupa pelatihan pembuatan lama pekon menggunakan wordpress dilaksanakan agar desa-desa dapat memanfaatkan laman tersebut untuk mempromosikan potensinya masing-masing. Keberhasilan dari pelatihan ini diamati dengan memberikan ujian awal dan akhir. Hasilnya menunjukkan bahwa pelatihan ini memberikan peningkatan pengetahuan para peserta mengenai pemanfaatan wordpress untuk membangun laman pekon yang informatif.]]>
<![CDATA[STACKING ENSEMBLE APPROACH IN STATISTICAL DOWNSCALING USING CMIP6-DCPP FOR RAINFALL ESTIMATION IN RIAU ]]>
<![CDATA[Mahkya, Dani Al]]>;
<![CDATA[Djuraidah, Anik]]>;
<![CDATA[Wigena, Aji Hamim]]>;
<![CDATA[Sartono, Bagus]]>
<![CDATA[ MEDIA STATISTIKA Vol 17, No 1 (2024): Media Statistika ]]>
Publisher : <![CDATA[Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.14710/medstat.17.1.1-12
<![CDATA[Rainfall modeling and prediction is one of the important things to do. Rainfall has an important relationship and role with various aspects of the environment. One phenomenon that can be associated with rainfall is forest and land fires. Riau is one of the provinces in Indonesia that has a high potential for forest and land fires. This is because Riau has a large area of peatland. One approach that can be used to estimate rainfall is statistical downscaling. The concept of this approach is to form a functional relationship between global and local data. This research uses CMIP6-DCPP output data that will be used to estimate rainfall at 10 observation stations in Riau. The proposed model in this research is Stacking Ensemble with PC Regression and LASSO Regression in the base model and Multiple Linear Regression in the meta model. This research aims to determine the best CMIP6-DCPP model for estimating rainfall in Riau and increasing the accuracy of rainfall estimates using the Stacking Ensemble approach.]]>
