Articles

PERBANDINGAN METODE K–MEANS DAN SELF ORGANIZING MAP (STUDI KASUS: PENGELOMPOKAN KABUPATEN/KOTA DI JAWA TENGAH BERDASARKAN INDIKATOR INDEKS PEMBANGUNAN MANUSIA 2015) Kusumah, Rachmah Dewi; Warsito, Budi; Mukid, Moch. Abdul
Jurnal Gaussian Vol 6, No 3 (2017): Jurnal Gaussian
Publisher : Departemen Statistika FSM Undip

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (400.884 KB)

Abstract

Cluster analysis is a process of separating the objects into groups, so that the objects that belong to the same group are similar to each other and different from the other objects in another group. In this study used two method to classify data of  district / city in Central Java based on indicators of Human Development Index (HDI) 2015 are K-Means and Self Organizing Map (SOM) with the number of groups as much as two to seven. Furthermore, the results of both methods were compared using the Davies-Bouldin Index (DBI) values to determine which method is better. Based on the research that has been conducted found that the K-Means (K=4) method works better than SOM (K=2) to classify district / city in Central Java based on indicators of Human Development Index (HDI) as evidenced by the value of the Davies-Bouldin Index (DBI) on K-Means (K=4) of 0.786 is smaller than the value at SOM (K=2) Davies-Bouldin Index (DBI) which is equal to 0.893. Keywords: clustering, HDI, K-Means, SOM, DBI
PEMODELAN RETURN PORTOFOLIO SAHAM MENGGUNAKAN METODE GARCH ASIMETRIS Arifin, Muhammad; Tarno, Tarno; Warsito, Budi
Jurnal Gaussian Vol 6, No 1 (2017): Jurnal Gaussian
Publisher : Departemen Statistika FSM Undip

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (669.39 KB)

Abstract

Investment in stocks is an alternative for investors and companies to obtain external funding sources. In the investment world there is a strong relationship between risk and return (profit), if the risk is high then return will also be high. Risks can be minimized by performing stock portfolio. Stock is the time series data in the financial sector, which usually has a tendency to fluctuate rapidly from time to time so that variance of error is not constant. Time series model in accordance with these condition is Generalized Autoregressive Conditional Heteroscedasticity (GARCH). This research will apply asymmetric GARCH covering Exponential GARCH (EGARCH), Threshold GARCH (TGARCH), and Autoregressive Power ARCH (APARCH) in stock data Indocement Tunggal Tbk (INTP), Astra International Tbk (ASII), and Adaro Energy Tbk (ADRO) commencing from the date of March 1, 2013 until February 29, 2016 during an active day (Monday to Friday). The purpose of this research is to predict the value of the volatility of a portfolio of three assets stocks. The best models used for forecasting volatility in asset stocks which have asymmetric effect is ARIMA ([13],0,[2,3]) EGARCH (1,1) on a single asset data INTP, ARIMA ([2],0,[2,3]) EGARCH (1,1) on the 2 asset portfolio data ASII INTP, and ARIMA ([3],0,[2]) EGARCH (1,1) on the 3 asset portfolio data INTP-ASII-ADRO.Keywords: Stocks, Portfolio, Return, Volatility, Asymmetric GARCH.
IMPLEMENTASI ALGORITMA MODIFIED GUSTAFSON-KESSEL UNTUK CLUSTERING TWEETS PADA AKUN TWITTER LAZADA INDONESIA Putri, Ratna Kencana; Warsito, Budi; Mustafid, Mustafid
Jurnal Gaussian Vol 8, No 3 (2019): Jurnal Gaussian
Publisher : Departemen Statistika FSM Undip

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (717.172 KB)

Abstract

Online social media is a new kind of media which is steadily growing and has become publicly popular. Due to its ability to spread informations rapidly and its easiness to access for internet users, social media provides new alternative to conduct advertising and product segmentation. Twitter is one of the most favored social media with 19.5 million users in Indonesia to the date. In this research, the application of text mining to cluster tweets from the @LazadaID Twitter account is done using the Modified Gustafson-Kessel clustering algorithm. The clustering process is executed five times with the number of cluster starts from two to six cluster. The results of this research indicate that the optimum number of clusters formed based on the Partition Coefficient and Classification Entropy validation index are three clusters. Those three clusters are tweets containing electronic stuff offers, discounts, and prize quizes. Tweets with the most retweets and likes are prize quiz tweets. PT Lazada Indonesia could use this kind of tweet to conduct advertising on social media Twitter because the prize quiz tweets are liked by the @LazadaID Twitter account followers.Keywords: Twitter, advertising, Lazada Indonesia, Gustafson-Kessel Clustering algorithm, validation index
PEMODELAN KASUS KEMISKINAN DI JAWA TENGAH MENGGUNAKAN REGRESI NONPARAMETRIK METODE B-SPLINE Rahmawati, Anisa Septi; Ispriyanti, Dwi; Warsito, Budi
Jurnal Gaussian Vol 6, No 1 (2017): Jurnal Gaussian
Publisher : Departemen Statistika FSM Undip

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (577.596 KB)

Abstract

Poverty is one of the diseases in the economy, so it must be cured or at least reduced. According to BPS (2016), poor people are people who have an average expenditure per capita per month below the poverty line. The poverty line in Central Java in 2016 amounted to Rp 317 348, - per capita per month. In 2016, the average level of poverty in the Java Island, Central Java province placed as the second highest after DIY. Many factors are thought to affect the level of poverty. In this study, the predictor variables used are the rate of economic growth (X1), unemployment rate (X2), and education level above high school to (X3). This study aims to obtain a model of the relationship between the factors that affect poverty on the percentage of poor and calculate the predictions. The method used is B-spline nonparametric regression. Nonparametric approach are used if the function of previous data is unknown. The best B-spline model depends on the determination of the optimal knots point having a minimum Generalized Cross Validation (GCV). In this study, the best B-spline model obtained when the order of X1is 2, the order of X2 is 2, and the order of X3 is 2. The knots obtained in X1 at the point 4,51273, X2  at the point 3,60626, and X3 at point 11,4129 and 16,2481 with GCV value of 9,79353. Keywords: Poverty, Nonparametric Regression, B-Spline, Generalized Cross Validation
ANALISIS KOMODITAS UNGGULAN PERIKANAN BUDIDAYA PROVINSI JAWA TENGAH TAHUN 2012-2016 MENGGUNAKAN METODE LOCATION QUOTIENT DAN SHIFT SHARE Manullang, Dian Mariana L; Rusgiyono, Agus; Warsito, Budi
Jurnal Gaussian Vol 7, No 1 (2018): Jurnal Gaussian
Publisher : Departemen Statistika FSM Undip

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (564.844 KB)

Abstract

Condition of capture fisheries is currently stagnating, even tended to decline, which is indicated by the decrease of production in some fishery development areas in Indonesia. Aquaculture is one solution that can be done. Central Java Province is a province that has a large aquaculture potential, therefore of course Central Java province has leading commodities that become the sector of regional economic development. This research discusses about the potential location for the development of each leading commodities in Central Java Province as a recommendation related to the centre of fisheries production. Analytical methods in this research are Location Quotient (LQ) dan Shift share. It used to see how big these locations have a potential in the development of aquaculture production and to identify spatial autocorrelation in the amount of aquaculture production using Moran’s index. The analysis of LQ and shift share shows that each district has a different potential in the development of leading commodities production. The value of the Moran’s index obtained equal to -0.1381, that is in the range of -1 <I ≤ 0, indicating that the presence of spatial autocorrelation is negative but small because of near to zero. It can be concluded that there is no similarity of the values between the districts or indicate that amount of aquaculture production among the districts in Central Java are not correlated.Keywords: Leading Commodities, Location Quotient (LQ), Shift Share, Moran’s  Index
PEMODELAN GENERAL REGRESSION NEURAL NETWORK UNTUK PREDIKSI TINGKAT PENCEMARAN UDARA KOTA SEMARANG Warsito, Budi; Rusgiyono, Agus; Amirillah, M. Afif
MEDIA STATISTIKA Vol 1, No 1 (2008): Media Statistika
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (86.16 KB) | DOI: 10.14710/medstat.1.1.43-51

Abstract

This paper is discuss about General Regression Neural Network (GRNN) modelling to predict time series data, i.e. the air pollution rate in Semarang City comprises the floating dust, carbon monoxide (CO) and nitrogen monoxide (NO). The GRNN model have four processing layer that are input layer, pattern layer, summation layer and output layer. The input variable is determined by the ARIMA model. The result of GRNN modelling shows that the network have a good performance both at predict in sample and predict out of sample, that can be seen from the mean square error.   Keywords: GRNN, predict, air pollution  
PREDIKSI TERJANGKITNYA PENYAKIT JANTUNG DENGAN METODE LEARNING VECTOR QUANTIZATION Hidayati, Nurul; Warsito, Budi
MEDIA STATISTIKA Vol 3, No 1 (2010): Media Statistika
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (235.618 KB) | DOI: 10.14710/medstat.3.1.21-30

Abstract

Learning Vector Quantization (LVQ) is a method that train the competitives layer with supervised. The competitives layer will learn automatically to classify the input vector given. If some input vectors has the short distance then the input vector will be grouped into the same class. The LVQ method can be used to classify the data into some classes or categories. At this paper, the LVQ method will be applied to classify if someone is suffer potenciate of heart desease or not. The data that be trained are 268 data of heart desease patient from UCI (University of California at Irvine) with 10 variables that are factors influence that infected of heart desease. From some trials showed that the learning rate (α) = 0.25, decrease of learning rate (Decα) = 0.1, and the minimum learning rate (Minα) = 0.001 are values that give a good prediction with level of accuracy is about 66.79 %.   Keywords: Learning Vector Quantization, Classify, Heart Desease
PEMODELAN HYBRID ARIMA-ANFIS UNTUK DATA PRODUKSI TANAMAN HORTIKULTURA DI JAWA TENGAH Tarno, Tarno; Rusgiyono, Agus; Warsito, Budi; Sudarno, Sudarno; Ispriyanti, Dwi
MEDIA STATISTIKA Vol 11, No 1 (2018): Media Statistika
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (506.342 KB) | DOI: 10.14710/medstat.11.1.65-78

Abstract

The research purpose is modeling adaptive neuro fuzzy inference system (ANFIS) combined with autoregressive integrated moving average (ARIMA) for time series data. The main topic is application of Lagrange Multiplier (LM) test for input selection, determining the number of membership function and generating rules in ANFIS. Based on partial autocorrelation (PACF) plot, the lag inputs which are thought have an effect to data are evaluated by using LM-test. Procedure of LM test is applied to determine the optimal number of membership functions. Based on the result, a number of rule-bases are generated. The best model is applied for forecasting potato production data in Central Java. The case study of this research is modeling monthly data of potato production from January 2004 up to December 2016. From empirical study, ANFIS optimal was obtained with lag-1 and lag-11 as inputs with two membership functions and two fuzzy rules. The hybrid method based on ARIMA and ANFIS is also implemented. The result of the prediction with a hybrid method is compared to the ANFIS and ARIMA. Based on the value of Mean Absolute Percentage Error (MAPE), hybrid model ARIMA-ANFIS has a good performance as a model of ANFIS and ARIMA individually.Keywords: Time Series, Potato production, hybrid, ANFIS, ARIMA, LM-test
ANALYSIS OF THE NUMBER INFANT AND MATERNAL MORTALITY IN CENTRAL JAVA INDONESIA USING SPATIAL-POISSON REGRESSION Prahutama, Alan; Warsito, Budi; Mukid, Moch. Abdul
MEDIA STATISTIKA Vol 11, No 2 (2018): Media Statistika
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (322.832 KB) | DOI: 10.14710/medstat.11.2.135-145

Abstract

Maternal and infant mortality are one of the most dangerous problems of the community since it can profoundly affect the number and composition of the population. Currently, the government has been taking heed on the attempt of reducing the number of maternal and newborn mortality in Central Java which requires data and information entirely. Poisson regression is a nonlinear regression that is often used to model the relationship between response variables in the form of discrete data with predictor variables in the form of discrete or continuous data. In space analysis, GWPR is one of method in space modeling which can model regional-based regression. It is based on some factors including the number of health facilities, the number of medical personnel, the percentage of deliveries performed with non-medical assistance; the average age of a woman's first marriage; the average education level of married women; average amount of per capita household expenditure; percentage of village status; the average rate of exclusive breastfeeding; percentage of households that have clean water and the percentage of poor people. Based on the analysis, it is revealed that the determinants of maternal and infant mortality in Central Java using Poisson and GWPR models, among others are the number of health facilities, the number of medical personnel, the average number of per capita household expenditure and the percentage of the poor. In the maternal and infant mortality model, the AIC value of GWPR model produces better modeling than Poisson regression. Keywords: Maternal and Infant mortality, Poisson, GWPR
PREDIKSI CURAH HUJAN KOTA SEMARANG DENGAN FEEDFORWARD NEURAL NETWORK MENGGUNAKAN ALGORITMA QUASI NEWTON BFGS DAN LEVENBERG-MARQUARDT Warsito, Budi; Sumiyati, Sri
Jurnal Presipitasi : Media Komunikasi dan Pengembangan Teknik Lingkungan Vol 3, No 2 (2007): Vol 3, No 2 (2007)
Publisher : Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (120.768 KB) | DOI: 10.14710/presipitasi.v3i2.46-52

Abstract

This paper study the rainfall prediction at Semarang City as time series data with Feed Forward Neural Network  (FFNN)  model.  The  learning  algorithm  that  be  used  are  the  Quasi  Newton BFGS and Levenberg-Marquardt algorithm. The input unit is determined based on the best of ARIMA model. The computation is done with use  Matlab 7.1 program with 1000 epoch, five unit of hidden layer, 100 replication  and use  input at lag  variabel  1,  12  and 13, respectively. The result shows that the prediction is good in relatively, where Quasi Newton BFGS algorithm result the Mean Square Error (MSE) that more accurate.