Claim Missing Document
Check
Articles

Found 35 Documents
Search

Accuracy of the Moving Averages and Deseasonalizing Methods for Trend, Cyclical and Seasonal Data Forecasting Saragih, Yoga Fromega; Darnius, Open
JMEA : Journal of Mathematics Education and Application Vol 2, No 3 (2023): Oktober
Publisher : JMEA : Journal of Mathematics Education and Application

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30596/jmea.v2i3.13735

Abstract

Forecasting or forecasting is an attempt to predict future conditions based on past state data. Moving Averages or moving average is a forecasting method that calculates the average value of a time series and then uses it to estimate the value in the next period. Deseasonalizing is part of the decomposition method which is included in the time series method. In this study, the Moving Average method and the Deseasonalizing method were used. The use of these two forecasting methods is to determine the accuracy of the forecasting method which is more accurate and close to the Mean Absolute Error (MAE) and Mean Squared Error (MSE) values. In this study the procedures used were problem identification, problem formulation, observation, data analysis and conclusion. The data taken in this study is data that contains trend, cyclical, and seasonal. For data containing trends on the moving averages method 15245.28 and 1430419308, for the Deseasonalizing method 28121.9504 and 1204814887. For Cyclical data on the Moving Averages method 4454.314465 and 28200197.22 for the Deseasonalizing method 13357.71283 and 254833253.4. For Seasonal data on Moving Averages 126.3839286 and 25479.38393 for the Deseasonalizing method 244.9971767 and 75372.32397. And for data containing these three patterns in the Moving Averages method 193.5385 and 65781.02 for the Deseasonalizing method 901.9566 and 1351418. From these results it can be concluded that the most effective trend data is the Deseasonalizing method, for Seasonal data the most effective method is the Moving Averages method, and for Cyclical Data the most effective method is the Moving Averages. Meanwhile, for data containing the three data patterns is the Moving Averages method.
REGULARISASI REGRESI LINIER BERGANDA PADA DATA BERDIMENSI TINGGI UNTUK MENGATASI EFEK MULTIKOLINEARITAS Nasution, Muhammat Rayyan; Sutarman, S; Darnius, Open; Rosmaini, Elly
MES: Journal of Mathematics Education and Science Vol 10, No 1 (2024): Edisi Oktober
Publisher : Universitas Islam Sumatera Utara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30743/mes.v10i1.9469

Abstract

Penelitian ini membahas model regresi linier berganda yang diberikan regularisasi dalam kasus data berdimensi tinggi (? ≫ ?), bertujuan untuk mengatasi efek multikolinearitas yang terdiri dari efek singularitas dan kualitas model yang buruk. Dalam penelitian ini mengembangkan model regresi linier berganda dengan menambahkan parameter penalti pada fungsi tujuan. Adapun data yang digunakan adalah data primer yang dibangkitkan dengan bahasa pemrograman python dengan tiga skenario sesuai dari penelitian sebelumnya. Metode yang digunakan yaitu Ordinary Least Squared (OLS), Least Absolute Shrinkage and Selection Operator (LASSO) dan Ridge dalam mengestimasi parameter model regresi. Mean Squared Error (MSE) digunakan sebagai metrik evaluasi kinerja model yang dibangun. Dari hasil simulasi yang dilakukan, diperoleh bahwa metode LASSO memberikan kualitas model terbaik dengan memberikan nilai MSE terendah dibandingkan model lainnya.
Analysis Of Mutual Fund Performance In Indonesia Using Robust Regression Doloksaribu, Arsella F; br Tarigan, Enita Dewi; Syahmarani, Aghni; Darnius, Open; Hasibuan, Citra Dewi
MES: Journal of Mathematics Education and Science Vol 10, No 1 (2024): Edisi Oktober
Publisher : Universitas Islam Sumatera Utara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30743/mes.v10i1.9595

Abstract

The Least Squares Method (MKT) is the best linear regression estimator if the classical assumption tests are met. However, if there are outliers in the data, this approach can provide inaccurate prediction results. Therefore, the purpose of this research is to overcome the inappropriate model due to outliers in variables that affect mutual fund performance in Indonesia. The variables in this study are risk level, inflation, fund size, turnover ratio, and cash flow. Robust regression is an approach designed to provide more accurate estimates without discarding observational data that indicates outliers. One of the estimation approaches is Least Trimmed Squares (LTS). This estimation minimizes the sum of squared residuals from h observations that are not considered outliers. The results showed that there were 10 outliers in the data, and the risk level variable had no effect on mutual fund performance, while the inflation, fund size, turnover ratio, and cash flow variables had a significant effect. So by comparing the  value and residual standard error of the two methods, it is found that the  value in the LTS method is greater than the MKT method, namely 0.589 0.273, and the residual standard error in the LTS method is smaller than the MKT method, namely 9.73 48.59. Therefore, it can be concluded that the Least Trimmed Squares method provides better estimation results and is more effective for handling outliers than the MKT method. 
KLASIFIKASI PROVINSI DI INDONESIA BERDASARKAN LUAS PENGUSAHAAN DAN PRODUKTIVITAS TANAMAN KELAPA SAWIT MENGGUNAKAN ANALISA KLASTER Siahaan, Megawati; Darnius, Open
Agro Estate Vol 7 No 2 (2023): Desember 2023
Publisher : Institut Teknologi Sawit Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47199/jae.v7i2.164

Abstract

Luas areal perkebunan kelapa sawit di Indonesia semakin meningkat terus dengan pengusahaan oleh perkebunan besar negara, perkebunan besar swasta dan perkebunan rakyat. Produktivitas tanaman kelapa sawit bervariasi di perusahaan-perusahaan tersebut demikian juga dengan distribusinya di berbagai provinsi. Kelapa sawit ditanam dan tumbuh di berbagai provinsi di Indonesia, perlu dikelompokkan bagaimana pengelolaan kelapa sawit di masing-masing provinsi untuk mengetahui variasi pengelolaan kelapa sawit antar provinsi dan diperoleh upaya memperbaiki produktivitas berdasarkan pengelompokan tersebut. Data yang akan dianalisis merupakan data sekunder yang dikumpulkan dengan studi pustaka. Data dianalisa dengan analisis klaster menggunakan program IBM SPSS 22, dengan variabel terikat yang digunakan adalah provinsi di Indonesia sedangkan variabel bebasnya adalah luas areal perkebunan kelapa sawit yang pengusahaannya oleh perkebunan besar negara, perkebunan besar swasta dan perkebunan rakyat serta produktivitas kelapa sawit di masing-masing perkebunan tersebut sehingga ada total 6 variabel. Hasil dari analisis menunjukkan bahwa diperoleh 5 klaster provinsi di Indonesia yang mengelola kelapa sawit yaitu provinsi yang terbaik adalah Riau, Provinsi yang baik adalah Jambi, Provinsi yang cukup baik adalah Kalimatantan Barat, provinsi yang kurang baik adalah Sumatera Utara, Sumatera Selatan, Kalimantan Selatan, Provinsi yang paling tidak baik adalah Aceh, Sumatera Barat, Bengkulu, Lampung, Jawa Barat, Banten, Sulawesi Selatan, Sulawesi Tenggara.
Analysis of Internal Community Factors in Participating in the Family Planning Program in Medan Johor District Using the Logistic Regression Method Hardiyanti, Siska; Darnius, Open; Sitepu, Israil; S, Benar
Journal of Research in Mathematics Trends and Technology Vol. 5 No. 1 (2023): Journal of Research in Mathematics Trends and Technology (JoRMTT)
Publisher : Talenta Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32734/jormtt.v5i1.16026

Abstract

Regression is a statistical analysis method to describe the relationship between the dependent variable which has two or more categories with one or more independent variables on a categorical or continuous scale. The purpose of this study is to apply the logistic regression method to see how much influence the Internal Factor Analysis of the Community has in participating in the Family Planning Program in Medan Johor District. As is known, according to data from the BPS for North Sumatra Province, Medan Johor is included in a slum sub-district because it is caused by the dense population in this sub-district which is not in accordance with its area. The total population in Medan Johor District currently reaches 137,367 people, with an area of 16.96 km2and a population density of 8,099 people/km2. So that in this study, it was found that the level of education, age and occupation did not have a significant effect on opinion, which had a significant effect on opinion with a Sig value of 0.020 <0.05, namely the community's knowledge of the family planning program
MODEL HIBRIDA AUTOREGRESSIVE INTEGRATED MOVING AVERAGE (ARIMA) DAN FUZZY TIME SERIES (FTS) UNTUK PERAMALAN PRODUKSI KELAPA SAWIT PT. PERKEBUNAN NUSANTARA II Mayanti, Utari Sri; Darnius, Open; Sitepu, Israil
CARTESIUS : Jurnal Pendidikan Matematika Vol 6 No.1 Tahun 2023
Publisher : Unika Santo Thomas Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Model autoregressive integrated moving average (ARIMA) merupakan model yang secara penuh mengabaikan independen varibel dalam pembuatan peramalan. Untuk menghasilkan peramalan dalam jangka pendek yang akurat metode ARIMA menggunakan nilai masa lalu dan sekarang dari variabel dependen. Peramalan dengan menggunakan model ARIMA masih memiliki kekurangan dengan nilai kesalahan pengukuran yang cukup besar selain itu pada beberapa data time series terkadang mengandung pola linier maupun nonlinier sekaligus di dalamnya, maka diperlukan penggabungan model lain untuk meramalkan data non linier yang efektif seperti model fuzzy. Fuzzy Time Series (FTS) merupakan konsep yang digunakan untuk meramalkan masalah, dimana data aktual diubah menjadi nilai-nilai linguistik. Dengan Hibrida ARIMA dan FTS ditemukan model terbaik dengan Pemodelan Hibrida ARIMA (6,1,4) dan FTS pembobot Cheng dengan nilai RMSE terkecil yaitu sebesar 0,10615.
MODEL HIBRIDA AUTOREGRESSIVE INTEGRATED MOVING AVERAGE (ARIMA) DAN FUZZY TIME SERIES (FTS) UNTUK PERAMALAN PRODUKSI KELAPA SAWIT PT. PERKEBUNAN NUSANTARA II Mayanti, Utari Sri; Darnius, Open; Sitepu, Israil
CARTESIUS : Jurnal Pendidikan Matematika Vol 6 No.1 Tahun 2023
Publisher : Unika Santo Thomas Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Model autoregressive integrated moving average (ARIMA) merupakan model yang secara penuh mengabaikan independen varibel dalam pembuatan peramalan. Untuk menghasilkan peramalan dalam jangka pendek yang akurat metode ARIMA menggunakan nilai masa lalu dan sekarang dari variabel dependen. Peramalan dengan menggunakan model ARIMA masih memiliki kekurangan dengan nilai kesalahan pengukuran yang cukup besar selain itu pada beberapa data time series terkadang mengandung pola linier maupun nonlinier sekaligus di dalamnya, maka diperlukan penggabungan model lain untuk meramalkan data non linier yang efektif seperti model fuzzy. Fuzzy Time Series (FTS) merupakan konsep yang digunakan untuk meramalkan masalah, dimana data aktual diubah menjadi nilai-nilai linguistik. Dengan Hibrida ARIMA dan FTS ditemukan model terbaik dengan Pemodelan Hibrida ARIMA (6,1,4) dan FTS pembobot Cheng dengan nilai RMSE terkecil yaitu sebesar 0,10615.
Adjusting Anomalies in International Tourist Arrivals to North Sumatra During the Peak COVID-19 Period (April 2020 to June 2022) to Enhance the Validity of Time Series Modeling Eddy, Thaswin; Open Darnius
Journal of Research in Mathematics Trends and Technology Vol. 7 No. 2 (2025): Journal of Research in Mathematics Trends and Technology (JoRMTT)
Publisher : Talenta Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32734/jormtt.v7i2.21718

Abstract

The feasibility of time series modeling is significantly influenced by both the availability and the structural patterns of the data. Regular and continuous data collection over time is essential for constructing reliable time series models, particularly for forecasting purposes. Generally, a minimum of 50 time series data points is considered ideal to ensure the robustness and predictive power of such models. However, the presence of extreme fluctuations—such as sharp spikes or drops—can severely affect the integrity of the model. In the context of international tourist arrivals to North Sumatra during the peak period of the COVID-19 pandemic (April 2020 to June 2022), substantial data anomalies were observed. The results of modifying these anomalies indicate that increasing the number of adjusted data points during this period leads to a greater number of feasible time series models suitable for predictive analysis.
Penaksiran Parameter Pada Distribusi Erlang Berdasarkan Metode Maksimum Likelihood Dengan Menggunakan Algoritma Newton Raphson Dan Fisher Scoring Meili Yanti; Open Darnius
JURNAL RISET RUMPUN MATEMATIKA DAN ILMU PENGETAHUAN ALAM Vol. 2 No. 1 (2023): April : Jurnal Riset Rumpun Matematika dan Ilmu Pengetahuan Alam
Publisher : Pusat riset dan Inovasi Nasional

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55606/jurrimipa.v2i1.720

Abstract

The Erlang distribution is a special case of the Gamma distribution with the k shape parameter and the λ rate parameter. In this study, the parameter estimation of the Erlang distribution was carried out using the Maximum Likelihood method. In maximizing the function, an implicit and non-linear form is obtained, then it is solved using the Newton Raphson algorithm. Apart from Newton Raphon, the estimation of parameters was also carried out using the Fisher Scoring algorithm. The Fisher Scoring algorithm is similar to the Newton Raphson algorithm, the difference is that Fisher Scoring uses an matrix information. The result of parameter estimation in Erlang distribution using Newton Raphson algorithm which is applied to outgoing telephone call data that generated by Matlab R2010a software cannot be done simultaneously. Therefore, the parameter assessment is carried out on the k parameter first, then followed by the λ parameter estimation and the parameter and = 0.6886812 are obtained. Meanwhile, the parameter estimation using the Fisher Scoring algorithm produces an equation that is not different from the Newton Raphson algorithm
Analisis Faktor-faktor yang Mempengaruhi Tingkat Kemiskinan di Sumatera Utara Menggunakan Structural Equation Modeling Yessica Thania Silaban; Elly Rosmaini; Open Darnius; Asima Manurung
Konstanta : Jurnal Matematika dan Ilmu Pengetahuan Alam Vol. 2 No. 2 (2024): Juni : Jurnal Matematika dan Ilmu Pengetahuan Alam
Publisher : International Forum of Researchers and Lecturers

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59581/konstanta.v2i2.3220

Abstract

Poverty is a fundamental problem that is the center of attention of the government in any country. The aim of this research is to analyze the influence of education, unemployment, the Covid-19 pandemic and the human development index on poverty in North Sumatra. The data used in this research is secondary data based on time series available on the official website of the Central Statistics Agency in North Sumatra. This research data processing uses the help of SmartPLS 3 software. The research analysis used in this research is Structural Equation Modeling which shows that the variables education, unemployment, the Covid 19 pandemic and the human development index are exogenous variables and poverty is an endogenous variable. The research results obtained an R-Square value of 0.511 or 51.1%. The large value of the coefficient of determination shows that the independent variables consisting of education, unemployment, the Covid-19 pandemic and the human development index are able to explain the dependent variable, namely the poverty percentage of 51.1%. Meanwhile, the remaining 48.9% is explained by other variables not included in this research model. The human development index variable has a negative and significant effect on poverty of 0.678. For the Covid 19 Pandemic variable, it has a negative and significant effect on poverty of 0.267.