Articles
84
Documents
ESTIMASI REGRESI NON PARAMETRIK DENGAN METODE WAVELET SHRINKAGE NEURAL NETWORK PADA MODEL RANCANGAN TETAP

MEDIA STATISTIKA Vol 2, No 1 (2009): Media Statistika
Publisher : MEDIA STATISTIKA

Show Abstract | Original Source | Check in Google Scholar | Full PDF (371.536 KB)

Abstract

If X is a predictor variable and Y is a response  variable of following model Y = g(X) +e with function g is a regression which not yet been known and e is an independent random variable with mean 0 and variant . The function of g can be estimated by parametric and nonparametric approach. In this paper, g is estimated by nonparametric approach that is named wavelet shrinkage neural network  method. At this method, the smoothly function estimation is depending on shrinkage parameter’s that are threshold value and level of wavelet that be used. It also depending on the number of neuron in the hidden layer and the number of epoch that be used in feed forward neural network. Therefore, it is required to be select the optimal value of threshold, level of wavelet, the number of neuron and the number of epoch to determine optimal function estimation.   Keywords: Nonparametric Regression, Wavelet Shrinkage Neural Network

PEMILIHAN VARIABEL PADA MODEL GEOGRAPHICALLY WEIGHTED REGRESSION

MEDIA STATISTIKA Vol 4, No 2 (2011): Media Statistika
Publisher : Jurusan Statistika FSM Undip

Show Abstract | Original Source | Check in Google Scholar | Full PDF (439.595 KB)

Abstract

Regression analysis is a statistical analysis that aims to model the relationship between response variable with some predictor variables. Geographically Weighted Regression (GWR) is statistical method used for analyzed the spatial data in local form of regression. One of the problems in GWR is how to choose the significant variables. The number of predictor variables will allow the violation of assumptions about the absence of multicollinearity in the data. Therefore, this needs a method to reduce some of the predictor variables which not significant to the response variable. This paper will discuss how to select significant variables by stepwise method. This method is a combination of forward selection method and the backward elimination method. Keywords:   Geographically Weighted Regression, Backward Elimination, Forward Selection, Stepwise Method

ESTIMASI PARAMETER REGRESI LOGISTIK MULTINOMIAL DENGAN METODE BAYES

Jurnal Gaussian Vol 2, No 1 (2013): Wisuda Periode Januari 2013
Publisher : Jurusan Statistika UNDIP

Show Abstract | Original Source | Check in Google Scholar

Abstract

Multinomial logistic regression is a logistic regression where the dependent variable is polychotomous is dependent variable value of more than two categories. Multinomial logistic regression parameter estimation usually use classical method that is based only on current information obtained from the sample without taking into account the initial information of logistic regression parameters. If have early information  about parameter is prior distribution, the parameter estimation can use Bayes method. Bayesian methods combine information on the sample with prior distribution of information, and the results are expressed in the posterior distribution. If posterior distribution can not be derived analytically so approximated using Markov Chain Monte Carlo (MCMC) algorithm especially Metropolis-Hastings algorithm. This algorithm uses acceptance and rejection mechanism to generate a sequence of random samples. Keyword: Multinomial Logistic Regression, Bayes Method, Markov Chain Monte Carlo algorithm (MCMC), Metropolis-Hastings algorithm.

PEMODELAN LAJU INFLASI DI PROVINSI JAWA TENGAH MENGGUNAKAN REGRESI DATA PANEL

Jurnal Gaussian Vol 2, No 4 (2013): Wisuda Periode Oktober 2013
Publisher : Jurusan Statistika UNDIP

Show Abstract | Original Source | Check in Google Scholar

Abstract

Panel regression is a regression which is a combination of cross section and time series. To estimate the panel regression there are 3 approaches, the common effect model (CEM), the fixed effect model (FEM) and the random effect model (REM). In the CEM, the parameters were estimated using the Ordinary Least Square (OLS). In the FEM, the parameters estimated by OLS through the addition of dummy variables. At REM, error is assumed random and estimated by the method of Generalized Least Square (GLS). This study aims to analyze the factors that influence inflation in the Central Java province using panel regression. Based on test result of panel regression, the appropriate model is the CEM. The parameters of model are estimated by using OLS the cross section weights. The model show that the Consumer Price Index (CPI), Minimum Salary of City/Regency (MSCR) and the economic growth significantly effect on percentage of inflation in Central Java Province.

PENENTUAN FAKTOR PRIORITAS MAHASISWA DALAM MEMILIH TELEPON SELULER MERK BLACKBERRY DENGAN FUZZY AHP

Jurnal Gaussian Vol 1, No 1 (2012): Wisuda Periode Oktober 2012
Publisher : Jurusan Statistika UNDIP

Show Abstract | Original Source | Check in Google Scholar

Abstract

This study aims to determine the priority factor Diponegoro University students in choosing a BlackBerry mobile phone brands. Consumer or buyer is often confused in making the decision to buy a product because of the many factors that affect the choices available. From the method of Analytic Hierarchy Process (AHP) was found to be too subjective assessment of uncertainty for qualitative data. The problems above can be solved by the method of Fuzzy Analytic Hierarchy Process (FAHP), which uses the interval so that the assessment of qualitative data can provide a more objective assessment. The criteria used to be in this research are quality, price, design, and service. The data were taken by spreading questionnaires. From the answer of respondent, calculation of ratio was performed with a consistency ratio (CR). If CR<0.10 it means the answer of respondent is consisten and can be used for Fuzzy AHP. Based on the result of research, it could be concluded that quality was the top priority with 0.278 priority weights, then the service with 0.254 priority weights, design with 0.240 priority weight, and price with 0.228 priority weights.

APLIKASI MODEL REGRESI SPASIAL UNTUK PEMODELAN ANGKA PARTISIPASI MURNI JENJANG PENDIDIKAN SMA SEDERAJAT DI PROVINSI JAWA TENGAH

Jurnal Gaussian Vol 2, No 4 (2013): Wisuda Periode Oktober 2013
Publisher : Jurusan Statistika UNDIP

Show Abstract | Original Source | Check in Google Scholar

Abstract

Net Enrollment Ratio (NER) is an  instrument to measure education rate. But NER rate of Senior High School in Central Java Province is only 47,34 %. This study discuss about regression model of factors which influence NER of Senior High School for Central Java province considering  spatial effects for each regency  in Central Java province. The examination of spatial effects shows that there is spatial dependence in response variable so this study is developed by using Spatial Autoregressive Model (SAR). The methods for estimating the parameter are   Ordinary Least Square and Maximum Likelihood Estimation. The result of this study shows that the average number of household members has significant spatial effect for NER rate of Senior High School in Central Java Province. From the comparison AIC value, it was found that SAR model is better to analyze NER rate of Senior High School in Central Java province than classic one.

ANALISIS FAKTOR-FAKTOR YANG MEMPENGARUHI BAYI BERAT LAHIR RENDAH DENGAN MODEL REGRESI LOGISTIK BINER MENGGUNAKAN METODE BAYES (Studi Kasus di Rumah Sakit Umum Daerah Kota Semarang)

Jurnal Gaussian Vol 1, No 1 (2012): Wisuda Periode Oktober 2012
Publisher : Jurusan Statistika UNDIP

Show Abstract | Original Source | Check in Google Scholar

Abstract

This study aims to elucidate several factors which affect low-birth-weight (LBW) infants in Semarang General Hospital (RSUD) in the period from July to December 2011. With regard to the MilleniumDevelopment Goals’s targets, which are predominantly intended to reduce the child mortality rate, serious investigations are highly needed to identify the factors that determine the rate of babies born with the low-birth-weight category. This problem can be solved with the Binary Logistic Regression model,using the Bayesian method. The Bayesian method is one of the parameter estimation technique which employ prior value as initial knowledge. The conducted research is to argue that both factors of age and the maternal hemoglobin level considerably give influence on LBW birth. Based on the research analysis, it is extremely recommended that mother to be pays much attention not to be pregnant at relatively young age and maintain the secure level of hemoglobin during pregnancy. 

OPTIMASI WAKTU EFEKTIF APLIKASI HERBISIDA PADA TANAMAN KELAPA SAWIT (ELAEIS GUINEENSIS JACQ.) DENGAN FUNGSI ESTIMASI DENSITAS KERNEL (Studi Kasus di Perkebunan Sawit PT SMART Tbk, Libo Estate, Riau)

Jurnal Gaussian Vol 1, No 1 (2012): Wisuda Periode Oktober 2012
Publisher : Jurusan Statistika UNDIP

Show Abstract | Original Source | Check in Google Scholar

Abstract

Palm oil agribusiness is one of potential source to accelerate economic growth in Indonesia. Palm oil is the raw material to produce CPO (Crude Palm Oil) which is source of vegetable oil that is needed by all people. This research used a combination of 16 treatments of type and dose  of herbicide on oil palm trees. Purposes of this research are to determine the optimal timing of herbicide applications and determine the treatment that maximizes efficacy of weed. Optimal timing of herbicide applications to the palm trees is determined through the largest mean of bootstrap resample and plot of kernel epanechnikov density estimation. Optimal treatment is determined through the largest mean of bootstrap resample, the smallest variance resample, the smallest range of bootstrap percentile confidency interval, and coverage probability that close to 1-α. Result obtained is the optimal timing of herbicide applications to oil palm trees is 8 weeks after applications. And optimal treatment is Tricalon 318 EC at a dose of 1500 cc.

FAKTOR-FAKTOR YANG MEMPENGARUHI STATUS KELULUSAN BERDASARKAN JALUR MASUK MAHASISWA DENGAN MODEL REGRESI LOGISTIK BINER BIVARIAT (Studi Kasus Mahasiswa FSM Universitas Diponegoro)

Jurnal Gaussian Vol 2, No 4 (2013): Wisuda Periode Oktober 2013
Publisher : Jurusan Statistika UNDIP

Show Abstract | Original Source | Check in Google Scholar

Abstract

The acceptance of college students in public universities are divided into two ways, the National Selection of Public University Entrance by invitation and the National Selection of Public University Entrance by non invitation. The National Selection of Public University Entrance by invitation is a way to get candidate students from The Senior High Schools that have good achievement, where as the other one open wider access. Nevertheless, the college students who enter through the invitation or non invitation, they don’t necessarily have a better academic achievement or worse than each other. After through the learning process in college, the success of the students are marked with their academic achievement that shown by the index of academic achievement, that if they pass expressed by the status of graduation, cumlaude or not cumlaude. To find out the factors that affect the status of student graduation based on the entrance, the regretion model that can be used is bivariate biner logistic regretion, because it consist of two response variable, the status of graduation and the entrance of the college students. Maximum likelihood estimation is used to estimate the parameter model. To examine the significance of the parameter use Likelihood ratio test and Wald test. Major option and live adress are the significance variables that affect the status of graduation based on the entrance of the college student from predictor variable partially test of school report grades, national test grades, major option, live adress, study method, live cost, students’ relationship with friends and family,and study motivation. Whole test and individual test indicate that major option variable affect the status of graduation based on the entrance significantly.

ESTIMASI KANDUNGAN HASIL TAMBANG MENGGUNAKAN ORDINARY INDICATOR KRIGING

Jurnal Gaussian Vol 2, No 1 (2013): Wisuda Periode Januari 2013
Publisher : Jurusan Statistika UNDIP

Show Abstract | Original Source | Check in Google Scholar

Abstract

Kriging is a geostatistical analysis of the data used to estimate the value that represents a no sample point based sample point in the surrounding by considering the spatial correlation in the data. Kriging is an interpolation method that generates unbiased predictions or estimations and has a minimum error. Indicator kriging is an estimation method that does not require the assumption of normality of data and can also be used to treat data that have a significant outlier. The indicator kriging that based on the principle of ordinary kriging also called ordinary indicator kriging. In this case study conducted Morowali estimated iron content in Central Sulawesi using ordinary indicator kriging method. The data used in the form of data coordinate point and iron content. The results obtained are presented probability value locations that fall within the zone of potential and non potential with the value the error variance. Based on the analysis to obtain a plot depicting the location of the entry in the zones of potential iron mine on the abscissa coordinate (7150–7210), the ordinate (54180–54540), and the depth ranges (440–500) meters and also the coordinates of the abscissa (7710–8130), the ordinate (54800–54960), and depths ranging from (327–342) meters.

Co-Authors Abdul Hoyyi Agus Rusgiyono Ahmat Dhani Riau Bahtiyar Alan Prahutama Aldila Abid Awali Alfiyatun Rohmaniyah Alvita Rachma Devi Amanda Lucky Berlian Andriyani, Mega Fitria Arief Rachman Hakim Budi Warsito Cakra Kurniawan Catra Aditya Wisnu Aji Christa Monica, Christa Danang Chandra Pradana, Danang Chandra Dani Al Mahkya Darwanto Darwanto David Yuliandar Depy Veronica, Depy Dewi Setya Kusumawardani Di Asih I Maruddani Diah Safitri Dody Apriliawan Dwi Hasti Ratnasari Dwi Ispriyanti Erfan Sofha, Erfan Fiqria Devi Ariyani Firda Shintia Dewi Gera Rozalia, Gera Hanien Nia H Shega Hari Susanta Nugraha Ilham Indra Bakti Al-Irsyad, Ilham Indra Bakti Indra Satria, Indra Ira Puspita Sari Ishlahul Kamal, Ishlahul Johan Adi Wicaksana, Johan Adi Kadi Mey Ismail Khusnul Yeni Widiyanti Laily Nadhifah Lenti Agustina Lianasari Tambunan Lina Irawati Lutfia Septiningrum, Lutfia Mamuroh Mamuroh Maulana Taufan Permana Meilia Kusumawardani, Meilia Moch. Abdul Mukid Muh Najib Hilmi MUHAMMAD HARIS Muhammad Hilman Rizki Abdullah, Muhammad Hilman Rizki Muhammad Mujahid Muryanto Muryanto, Muryanto Mustafid Mustafid Novian Trianggara, Novian Novika Pratnyaningrum Nur Indah Yuli Astuti, Nur Indah Yuli Nurmalita Sari, Nurmalita Nursihan Nursihan, Nursihan Putri Aulia Netra, Putri Aulia Putri Aulia Wahyuningsih Ragil Saputra Rahafattri Ariya Fauzannissa, Rahafattri Ariya Rahma Kurnia Widyawati Rahma Nurfiani Pradita, Rahma Nurfiani Restu Dewi Kusumo Astuti Rezzy Eko Caraka Riama Oktaviani Samosir, Riama Oktaviani Rifki Adi Pamungkas, Rifki Adi Rima Nurlita Sari, Rima Nurlita Rita Rahmawati Rizky Amanda Rose Debora Julianisa, Rose Debora Rukun Santoso Safitri Daruyani Sari, Ajeng Arum Sarita Budiyani Purnamasari Sudarno Sudarno Sugito Sugito Suparti Suparti Syaifudin Karyadi, Syaifudin Tarno Tarno Tatik Widiharih Tiani Wahyu Utami Tika Dhiyani Mirawati Triastuti Wuryandari Vita Dwi Rachmawati Wawan Sugiarto, Wawan Wayaning Apsari Yuciana Wilandari Yudha Subakti, Yudha Zulfa Wahyu Mardika, Zulfa Wahyu