p-Index From 2015 - 2020
11.106
P-Index
Suparti Suparti
Universitas Diponegoro

Published : 162 Documents
Articles

## Title

### Found 11 Documents Search Journal : MEDIA STATISTIKA

PEMILIHAN THRESHOLD OPTIMAL PADA ESTIMATOR REGRESI WAVELET THRESHOLDING DENGAN METODE CROSS VALIDASI Suparti, Suparti; Tarno, Tarno; Hapsari, Paula Meilina Dwi
MEDIA STATISTIKA Vol 2, No 2 (2009): Media Statistika
Publisher : MEDIA STATISTIKA

#### Abstract

If x is a predictor variable and y is a response&nbsp; variable of &nbsp;the regression model y = f (x)+ &Icirc; with&nbsp; f is a regression function which not yet been known and &Icirc; is independent random variable with mean 0 and variance , hence function f can be estimated by parametric and nonparametric approach. In this paper function f is estimated with a nonparametric approach. Nonparametric approach that used is a wavelet shrinkage or a wavelet threshold method. In the function estimation with a wavelet threshold method, &nbsp;the value of&nbsp; threshold has &nbsp;the most important role to determine &nbsp;level of smoothing estimator. The small threshold give function estimation very no smoothly, while&nbsp; the big value of threshold give function estimation very smoothly. Therefore the optimal value of threshold should be selected to determine the optimal function estimation. One of the methods to determine the optimal value of threshold by minimize a cross validation function. The cross validation method that be used is two-fold cross validatiaon. In this cross validation, it compute the predicted value by using a half of data set. The original data set is split &nbsp;into two subsets of equal size : one containing only the even indexed data, and the other, the odd indexed data. The odd data will be used to predict the even data, and vice versa. Based on &nbsp;the result of data analysis, the optimal threshold with cross validation method is not uniq, but they give the &nbsp;uniq of wavelet thersholding regression estimation. &nbsp; Keywords : Nonparametric Regression, Wavelet Threshold Estimator, Cross Validation.
ANALISIS FAKTOR-FAKTOR YANG MEMPENGARUHI BANYAKNYA KLAIM ASURANSI KENDARAAAN BERMOTOR MENGGUNAKAN MODEL REGRESI ZERO-INFLATED POISSON (Studi Kasus di PT. Asuransi Sinar Mas Cabang Semarang Tahun 2010) Taufan, Muhammad; Suparti, Suparti; Rusgiyono, Agus
MEDIA STATISTIKA Vol 5, No 1 (2012): Media Statistika
Publisher : Jurusan Statistika FSM Undip

#### Abstract

Poisson regression is one of model that is often used to model the relationship between response variables in the form of discrete data with a set of predictor variables in the form of continuous, discrete, category, or mixture data. In Poisson regression assumes that the mean of the response variable equal to the variance (equidispersion). But in reality, sometimes found a condition called overdispersion, that the variance value is greater than the mean. One of the cause of overdispersion is excess zero in the response variable. One of model that can be used to overcome this overdispersion problem is Zero-Inflated Poisson (ZIP) regressionÂ  model. This model is applied on a case study of motor vehicle insurance in the branch of PT. Asuransi Sinar Mas in Semarang in 2010 to determine the effect of age of car and types of coverage to number of claims filed by the policyholder to the branch of PT. Asuransi Sinar Mas in Semarang. In this case, the occurrence of zeros due to many policyholders did not file a claim to the branch of PT. Asuransi Sinar Mas in Semarang. From the analytical result obtained the conclution that the age of car and types of coverage affect number of claims filed by the policyholder to the branch of PT. Asuransi Sinar Mas in Semarang in 2010. Â  Keywords: Poisson Regression, Overdispersion, Zero-Inflated Poisson (ZIP) Regression
PEMILIHAN PARAMETER THRESHOLD OPTIMAL DALAM ESTIMATOR REGRESI WAVELET THRESHOLDING DENGAN PROSEDUR FALSE DISCOVERY RATE (FDR) Suparti, Suparti; Tarno, Tarno; Haryono, Yon
MEDIA STATISTIKA Vol 1, No 1 (2008): Media Statistika
Publisher : Jurusan Statistika FSM Undip

#### Abstract

If X is predictor variable and Y is response&nbsp; variable of following model Y = f (X) +e with function f is regression which not yet been known and e is independent random variable with mean 0 and variant , hence function of f can estimate with parametric and nonparametric approach. At this paper estimate f with nonparametric approach. Nonparametric approach that used is wavelet shrinkage or wavelet thresholding method. At function estimation with method of wavelet thresholding, what most dominant determine level of smoothing estimator is value of threshold. The small threshold give function estimation very no smoothly, while&nbsp; the big value of threshold give function estimation very smoothly. Therefore require to be selected value of optimal threshold to determine optimal function estimation. &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &nbsp; One of the method to determine the value of optimal threshold is with procedure of False Discovery Rate ( FDR). In procedure of FDR, the optimal threshold determined by selection of level of significance. Smaller mount used significance progressively smoothly its . &nbsp; Keywords: Nonparametric regression, wavelet thresholding estimator, procedure of False Discovery Rate
ANALISIS DATA INFLASI DI INDONESIA MENGGUNAKAN MODEL REGRESI SPLINE Suparti, Suparti
MEDIA STATISTIKA Vol 6, No 1 (2013): Media Statistika
Publisher : Jurusan Statistika FSM Undip

#### Abstract

The inflation data is one of the financial time series data that has a high volatility, so if the data is modeled with parametric models (AR, MA and ARIMA), sometimes occur problems because there was an assumption that cannot be satisfied. The developed model of parametric to cope with the volatility of the data is the ARCH and GARCH models. This alternative parametric models still requires the normality assumption in the data that often cannot be satisfied by financial data. Then a nonparametric method that does not require strict assumptions as parametric methods is developed. This research aims to conduct a study in Indonesia inflation data modeling using nonparametric methods is spline regression model with truncated spline bases. Goodness of a spline regression model is determined by an orde and knots location . However, the knots location are more dominant in spline regression model. One way to get the optimal knots location are by minimizing the value of Generalized Cross Validation (GCV). By modeling the annual inflation data of Indonesia in December 2006 - December 2011, the inflation target in 2012 is 4.5% + 1% can be achieved while the inflation target in 2013 is 4.5% + 1% cannot be achieved, because that prediction in 2013 is 8.55%. It was caused by government policy to raise the price of basic electricity and the fuel prices in 2013. Keywords : Inflation, Spline Regression Model, Generalized Cross Validation.
ANALISIS DATA INFLASI DI INDONESIA PASCA KENAIKAN TDL DAN BBM TAHUN 2013 MENGGUNAKAN MODEL REGRESI KERNEL Suparti, Suparti
MEDIA STATISTIKA Vol 6, No 2 (2013): Media Statistika
Publisher : Jurusan Statistika FSM Undip

#### Abstract

The inflation data is one of the financial time series data that has a high volatility, so if the data is modeled with parametric models (AR, MA and ARIMA), sometimes occur problems because there was an assumption that cannot be satisfied. Then a nonparametric method that does not require strict assumptions as parametric methods is developed. This study aims to analyze inflation in Indonesia after the goverment raised the price of electricity basic and fuel price in 2013 using kernel regression models. This method was good for data modeling inflation in Indonesia before. The goodness of a kernel regression model is determined by the chosen kernel function and wide bandwidth used. However, the most dominant is the selection of the wide bandwidth. In this study, determination of the optimal bandwidth by minimizing the Generalized Cross Validation (GCV). By model the annual inflation data (Indonesia) December 2006 - December 2011, the inflation target in 2012 is (4,5 + 1 )% can be achieved both exactly and predictly, while the inflation target in 2013 is (4,5 + 1 )% cannot be achieved neither exactly nor predictly. The inflation target in 2013 canât be achieve because since the beginning of 2013, there was a government policy to raise the price of electricity and the middle of 2013, there was an increase in fuel prices. The prediction of Indonesia inflation in 2014 by Gauss kernel is 6,18%. Keywords: Inflation, Kernel Regression Models, Generalized Cross Validation
BIPLOT UNTUK MENGETAHUI KARAKTERISTIK KABUPATEN/KOTA DI JAWA TENGAH BERDASARKAN PRODUKSI BAWANG PUTIH, BAWANG MERAH, CABE BESAR DAN CABE RAWIT Safitri, Diah; Suparti, Suparti; Pratiwi, Esti; Estiningrum, Tyas
MEDIA STATISTIKA Vol 7, No 1 (2014): Media Statistika
Publisher : Jurusan Statistika FSM Undip

#### Abstract

Biplot is a graphical representation of a data matrix. Garlic, onions, chili, and thai pepper are important plant in Indonesia because most people in Indonesia especially in Central Java consume garlic, onions, chili, and thai pepper every day. In this research, districts in Central Java seen characteristics are based on the productions of garlic, onions, chili, and thai pepper using biplot. There are highly correlation between chili and thai pepper, which means districts that have highly productions of chili will also tend to have highly production of thai pepper. There are some districts have the production ofÂ  garlic, onions, chili, and thai pepper relatively low, and there are some of the city has zero production ofÂ  garlic, onions, chili, and thai pepper. Â  Keywords: Biplot, Production ofÂ  garlic, onions, chili, thai pepper
ANALISIS DATA INFLASI INDONESIA MENGGUNAKAN MODEL AUTOREGRESSIVE INTEGRATED MOVING AVERAGE (ARIMA) DENGAN PENAMBAHAN OUTLIER Suparti, Suparti
MEDIA STATISTIKA Vol 8, No 1 (2015): Media Statistika
Publisher : Jurusan Statistika FSM Undip

#### Abstract

The inflation data is one of the financial time series data which often has high volatility. It is caused by the presence of outliers in the data. Therefore, it is necessary to analyze forecasting that can make all the assumptions are fulled without having to ignore the presence of outliers. The aim of this study is analyzing the inflation data in Indonesia using ARIMA model with the outlier detection. By modeling annual inflation data in December 2006 to December 2013 there are two types of outlier that are additive outlier (AO) and level shift (LS) outlier. The results show that The ARIMA model with the addition of outlier are better than the ARIMA model without outlier. The ARIMA ([1.12], 1.0) model with the addition of 19 outliers meet to the all assumptions that are the significance parameters, normality, homoscedasticity, and independence of residuals as well as the smallest MSE value. Keywords: Inflation, ARIMA, Outlier, MSE
PEMODELAN VOLATILITAS UNTUK PENGHITUNGAN VALUE AT RISK (VaR) MENGGUNAKAN FEED FORWARD NEURAL NETWORK DAN ALGORITMA GENETIKA Yasin, Hasbi; Suparti, Suparti
MEDIA STATISTIKA Vol 7, No 2 (2014): Media Statistika
Publisher : Jurusan Statistika FSM Undip

#### Abstract

High fluctuations in stock returns is one problem that is considered by the investors. Therefore we need a model that is able to predict accurately the volatility of stock returns. One model that can be used is a model Generalized Autoregressive Conditional Heteroskedasticity (GARCH). This model can serve as a model input in the model Feed Forward Neural Network (FFNN) with Genetic Algorithms as a training algorithm, known as GA-Neuro-GARCH. This modeling is one of the alternatives in modeling the volatility of stock returns. This method is able to show a good performance in modeling the volatility of stock returns. The purpose of this study was to determine the stock return volatility models using a model GA-Neuro-GARCH on stock price data of PT. Indofood Sukses Makmur Tbk. The result shows that the determination of the input variables based on the ARIMA (1,0,1) -GARCH (1,1), so that the model used FFNN consists of 2 units of neurons in the input layer, 5 units of neurons in the hidden layer neuron layer and 1 unit in the output layer. then using a genetic algorithm with crossover probability value of 0.4, was obtained that the Mean Absolute Percentage Error (MAPE) of 0,0039%. Â  Keywords: FFNN, Genetic Algorithm, GARCH, Volatility
PEMODELAN DATA INFLASI INDONESIA PADA SEKTOR TRANSPORTASI, KOMUNIKASI, DAN JASA KEUANGAN MENGGUNAKAN METODE KERNEL DAN SPLINE Suparti, Suparti; Tarno, Tarno
MEDIA STATISTIKA Vol 8, No 2 (2015): Media Statistika
Publisher : Jurusan Statistika FSM Undip

#### Abstract

In this research, we study data modeling of Indonesian inflation in theÂ  transportation, communication and financial services sector using the kernel and spline models. Determination of the optimal models based on the smallest of GCVÂ  value and determination of the best model based on the smallest out sampels of Mean Square Error (MSE) value. By modeling the yoy (year on year) inflation data in Indonesia in the transportation, communication and financial services sector In January 2007 to January 2015, shows that the kernel modelÂ  using Gaussian kernel function obtained optimal model with a bandwidthÂ  0.24 and the optimal spline model with order 5 andÂ  4 points knots. Based on out sampels dataÂ  in February to August 2015, obtained out sampelsÂ  MSE value of the spline model is smaller than the kernel model. So that the spline model is better than the kernel modelÂ  to analyzeÂ  the inflation dataÂ  of transportation, communication and financial services sector.Keywords: Inflation, Transportation, Communication and Financial Services Sector, Kernel, Spline, GCV, MSE.
PEMODELAN REGRESI NONPARAMETRIK MENGGUNAKAN PENDEKATAN POLINOMIAL LOKAL PADA BEBAN LISTRIK DI KOTA SEMARANG Suparti, Suparti; Prahutama, Alan
MEDIA STATISTIKA Vol 9, No 2 (2016): Media Statistika
Publisher : Departemen Statistika FSM Undip

#### Abstract

Semarang is the provincial capital of Central Java, with infrastructure and economicâs growth was high. The phenomenon of power outages that occurred in Semarang, certainly disrupted economic development in Semarang. Large electrical energy consumed by industrial-scale consumers and households in the San Francisco area, monitored or recorded automatically and presented into a historical data load power consumption. Therefore, this study modeling the load power consumption at a time when not influenced by the use of electrical load (t-1)-th. Modeling using nonparametric regression approach with Local polynomial. In this study, the kernel used is a Gaussian kernel. In local polynomial modeling, determined optimum bandwidth. One of the optimum bandwidth determination using the Generalized Cross Validation (GCV). GCV values obtained amounted to 1425.726 with a minimum bandwidth of 394. Modelling generate local polynomial of order 2 with MSE value of 1408.672.Â Keywords: electrical load, local polinomial, gaussian kernel, GCV.