Sudarno Sudarno
Departemen Statistika, Fakultas Sains Dan Matematika, Universitas Diponegoro

Published : 38 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 38 Documents
Search

ANALISIS SIX SIGMA DENGAN DECISION ON BELIEF CHART PADA PRODUK HOT STRIP MILL Alifia Hanifah Mumtaz; Mustafid Mustafid; Sudarno Sudarno
Jurnal Gaussian Vol 10, No 1 (2021): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/j.gauss.v10i1.29951

Abstract

The Decision On Belief (DOB) control chart is a univariate control chart that was initially introduced as a solution to the problem of less than optimal control limits from the shewhart attribute diagram, especially the control chart C. The new scheme based on the DOB control chart is that the calculation step , can change the data which initially is not normally distributed into a normal distribution, then can diagnose quality control process errors. . defines belief or an assumption in the new observation vector  and . The aim of this research is to apply the DOB control chart to data that is not normally distributed, so that it becomes a normal distribution. The result of the DOB control chart shows that the value of . is between the BKA and BKB values, which indicates a statistically controlled process. In this study, using one specification limit, namely the upper specification limit (USL) given by the company, which is 15 percent of the average production. The capability index used is  for 3 sigma using the transformation result . Based on the sample data, the result shows that the  value is 0.40633 and the sigma level is 2.719, so it can be concluded that the Hot Strip Mill production process is still not capable and has not reached the level of three sigma.Keywords: Six Sigma, Decision On Belief, capability index, , DPMO, level sigma. 
PENGELOMPOKAN TWEETS PADA AKUN TWITTER TOKOPEDIA MENGGUNAKAN ALGORITMA DENSITY BASED SPATIAL CLUSTERING OF APPLICATIONS WITH NOISE Deanira Qinanty Alamsyah; Sudarno Sudarno; Puspita Kartikasari
Jurnal Gaussian Vol 11, No 1 (2022): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/j.gauss.v11i1.33992

Abstract

Social media has become a trend for Indonesian people to express opinions, socialize, and exchange ideas. Internet users in Indonesia in 2021 will reach 202.6 million, 84% of whom use the internet to access social media. Twitter is one of the popular social media in Indonesia. This phenomenon is an opportunity for companies to use Twitter as a marketing tool, one of which is a marketplace company in Indonesia, Tokopedia. This research is intended to cluster tweets uploaded by the @tokopedia Twitter account to find out the type of content that gets a lot of likes and retweets by followers of the @tokopedia Twitter account. Cluster formation is done by applying the Density-Based Spatial Clustering of Applications with Noise algorithm (DBSCAN). DBSCAN is a clustering algorithm based on density. The DBSCAN algorithm requires two parameters, namely the radius (Eps) and the minimum number of objects to form a cluster (MinObj). This research conducted several experiments with different Eps and MinObj parameters on 1.344 tweets that had gone through the stages of removing duplication, text preprocessing, and feature selection. The quality of the cluster formed is measured using the Silhouette Coefficient. Based on the highest average Silhouette Coefficient, the parameter values of Eps=5 and MinObj=3 with Silhouette Coefficient = 0.575 are determined as the best parameters that produce 2 clusters and 7 noise. The type of content that has the highest average number of likes and retweets is the WIB (Indonesian Shopping Time) campaign, so Tokopedia can use this type of content as a marketing tool on Twitter social media because this type of content is preferred by followers of the @tokopedia Twitter account. Keywords: Twitter, Tokopedia, Clustering, DBSCAN, Silhouette Coefficient
ANALISIS ARIMA DAN WAVELET UNTUK PERAMALAN HARGA CABAI MERAH BESAR DI JAWA TENGAH Chrisentia Widya Ardianti; Rukun Santoso; Sudarno Sudarno
Jurnal Gaussian Vol 9, No 3 (2020): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/j.gauss.v9i3.28906

Abstract

Time series is a type of data collected according to the sequence of times in a certain time span. Time series data can be used as a predictor of future conditions. Analysis of time series data, one of the ARIMA units, is a parametric method that requires an assumption to get valid results. Data stationarity is one of the factors that must be fulfilled. Wavelet is a non-parametric method that is able to represent time and frequency information simultaneously, so that it can analyze non-stationary data. This research presents forecasting the price of red chili in Central Java using ARIMA and wavelet with the approach of the Multiscale Autoregressive (MAR) model. The best model is the one with the smallest MSE value. The results showed that the ARIMA(0,1,1) model was said to be the best model with MSE = 2252142. However, because the assumption of normality is not fulfilled, an alternative process is done with wavelet. Wavelet approach results show that the MAR model Haar filter level (j) = 4 with MSE = 2175906 is better than Daubechies 4 filter 4 level (j) = 1 with MSE = 3999669. Therefore, the Haar wavelet is considered better in the time series analysis. Keyword : ARIMA, wavelet, MAR, forecasting, MSE
PENERAPAN METODE PENGENDALIAN KUALITAS MEWMA BERDASARKAN ARL DENGAN PENDEKATAN RANTAI MARKOV (Studi Kasus: Batik Semarang 16, Meteseh) Enggartya Andini; Sudarno Sudarno; Rita Rahmawati
Jurnal Gaussian Vol 10, No 1 (2021): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/j.gauss.v10i1.30939

Abstract

An industrial company requires quality control to maintain quality consistency from the production results so that it is able to compete with other companies in the world market. In the industrial sector, most processes are influenced by more than one quality characteristic. One tool that can be used to control more than one quality characteristic is the Multivariate Exponentially Weighted Moving Average (MEWMA) control chart. The graph is used to determine whether the process has been controlled or not, if the process is not yet controlled, the next analysis that can be used is to use the Average Run Length (ARL) with the Markov Chain approach. The markov chain is the chance of today's event is only influenced by yesterday's incident, in this case the chance of the incident in question is the incident in getting a sampel of data on the production process of batik cloth to get a product that is in accordance with the company standards. ARL is the average number of sample points drawn before a point indicates an uncontrollable state. In this study, 60 sample data were used which consisted of three quality characteristics, namely the length of the cloth, the width of the cloth, and the time of the fabric for the production of written batik in Batik Semarang 16 Meteseh. Based on the results and discussion that has been done, the MEWMA controller chart uses the λ weighting which is determined using trial and error. MEWMA control chart can not be said to be stable and controlled with λ = 0.6, after calculating, the value is obtained Upper Control Limit (BKA) of 11.3864 and Lower Control Limit (BKB) of 0. It is known that from 60 data samples there is a Tj2 value that comes out from the upper control limit (BKA) where the amount of 15.70871, which indicates the production process is not controlled statistically. Improvements to the MEWMA controller chart can be done based on the ARL with the Markov Chain approach. In this final project, the ARL value used is 200, the magnitude of the process shift is 1.7 and the r value is 0.28, where the value of r is a constant obtained on the r parameter graph. The optimal MEWMA control chart based on ARL with the Markov Chain approach can be said to be stable and controlled if there is no Tj2 value that is outside the upper control limit (BKA). The results of the MEWMA control chart based on the ARL with the Markov Chain approach show that the process is not statistically capable because the MCpm value is 0.516797 and the MCpmk value is 0.437807, the value indicates a process capability index value of less than 1. Keywords: Handmade batik, Multivariate Exponentially Weighted Moving Average (MEWMA), Average Run Length (ARL), Capability process.
OPTIMASI REGRESI LOGISTIK MENGGUNAKAN ALGORITMA GENETIKA UNTUK PEMODELAN FAKTOR-FAKTOR YANG MEMPENGARUHI PENGGOLONGAN KREDIT BANK (Studi Kasus: Debitur di PT BPR Gunung Lawu Klaten Periode Tahun 2017) Reno Penggalih Surya Wardhani; Sudarno Sudarno; Di Asih I Maruddani
Jurnal Gaussian Vol 8, No 4 (2019): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (673.647 KB) | DOI: 10.14710/j.gauss.v8i4.26751

Abstract

Credit is the greatest asset managed by banks and also the most dominant contributor to the bank’s income. But in its implementation, the provision of credit to the public is at risk for non-performing loans. For this reason, creditors try to minimize the occurrence of non-performing loans by predicting credit risk appropriately. In this study, modeling the factors that influence credit classification at PT BPR Gunung Lawu is useful for predicting the credit risk of prospective debtors. Modeling are done using logistic regression and genetic algorithms. Factors suspected of influencing credit classification include age, gender, marital status, education, home ownership, employment, net income, tenor, type of business, type of loan, type of loan interest, and loan size. Estimated model parameters obtained from logistic regression were optimized using genetic algorithms. The fitness function used is pseudo  or  and MSE. The best model is generated by modeling with genetic algorithms based on MSE fitness. The model produces the highest  value of 0.1958 and the lowest MSE value of 0.1648 with classification accuracy of 75.33%. Keywords: credit classification, logistic regression, genetic algorithms
IMPLEMENTASI MODEL ACCELERATED FAILURE TIME (AFT) BERDISTRIBUSI LOG-LOGISTIK PADA PASIEN PENYAKIT JANTUNG BAWAAN Dwi Nooriqfina; Sudarno Sudarno; Rukun Santoso
Jurnal Gaussian Vol 10, No 3 (2021): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/j.gauss.v10i3.32796

Abstract

Log-Logistic Accelerated Failure Time (AFT) model is survival analysis that is used when the survival time follows Log-Logistic distribution. Log-Logistic AFT model can be used to estimate survival time, survival function, and hazard function. Log-Logistic AFT model was formed by regressing covariates linierly against the log of survival time. Regression coefficients are estimated using maximum likelihood method. This study uses data from Atrial Septal Defect (ASD) patients, which is a congenital disease with a hole in the wall that separates the top of two chambers of the heart by using sensor type III. Survival time as the response variable, that is the time from patient was diagnosed with ASD until the first relapse and uses age, gender, treatment status (catheterization/surgery), defect size that is the size of the hole in the heart terrace, pulmonary hypertension status, and pain status as predictor variables. The result showed that variable gender, treatment status, defect size, pulmonary hypertension status, and pain status affect the first recurrence of ASD patients, so it is found that category of female, untreated patient, defect size ≥12mm, having pulmonary hypertension, having chest pain tend to have first recurrence sooner than the other category.
PENERAPAN IMPROVED GENERALIZED VARIANCE PADA PENGENDALIAN KUALITAS PAVING BLOCK SEGIENAM Nathasa Erdya Kristy; Mustafid Mustafid; Sudarno Sudarno
Jurnal Gaussian Vol 9, No 1 (2020): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (830.059 KB) | DOI: 10.14710/j.gauss.v9i1.27526

Abstract

In quality assurance of hexagonal paving block products, quality control is needed so the products that produced are in accordance with the specified standards. Quality control carried out involves two interconnected quality characteristics, that is thickness and weight of hexagonal paving blocks, so multivariate control chart is used. Improved Generalized Variance control chart is a tool used to control process variability in multivariate manner. Variability needs to be controlled because of in a production process, sometimes there are variabilities that caused by engine problems, operator errors, and deffect in raw materials that affect the process. The purpose of this study is to apply Improved Generalized Variance control chart in controlling the quality of hexagonal paving block products and calculating the capability of production process to meet the standards. Based on the assumption of multivariate normal distribution test, it can be seen that the data of quality characteristics of hexagonal paving blocks have multivariate distribution. While based on the correlation test between variables it can be concluded that the characteristics of the quality of thickness and weight correlate with each other. The result of the control using these control chart shows that the process is statistically in control. The results of process capability analysis show that the production process has been running according to the standard because the process capability index value is generated using a weighting of 0.5 for each quality characteristic that is 1.01517. Keywords: Paving Block, Quality Control, Variability, Improved Generalized Variance, Process Capability Analysis
PERAMALAN MENGGUNAKAN MODEL FEED FORWARD NEURAL NETWORK DENGAN ALGORITMA ADAPTIVE SIMULATED ANNEALING (Studi kasus: Harga minyak mentah dunia yang dipublikasikan oleh OPEC) Affan Hanafaie; Sugito Sugito; Sudarno Sudarno
Jurnal Gaussian Vol 7, No 4 (2018): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/j.gauss.v7i4.28865

Abstract

Today, crude oil trading industry is still an important industry in the world because it still has high fuel oil consumption. The crude oil prices tend to fluctuate causing the prediction of crude oil in the coming periods to be a challenge. Forecasting the price of crude oil can be done by various methods, one of them is ARIMA Box-Jenkins model with OLS method to estimate the parameter, but this method has several assumptions that must be met. As time goes by, many methods that discovered, one of them is artificial neural network which can combined with various parameter optimization methods such as Adaptive Simulated Annealing algorithm. Adaptive Simulated Annealing algorithm is an optimization method that inspired by the process of crystallization, the advantages of this algorithm has a running time faster than similar algorithms. The combination of artificial neural networks and Adaptive Simulated Annealing algorithms can be used to model the historical data without requiring assumptions in the analysis. Based on the analysis on this research, the best model is obtained FFNN 2-5-1 with MAPE value of 1.0042%. Keywords: neural network, Adaptive Simulated Annealing, crude oil.
ANALISIS ANTREAN BUS NONPATAS JALUR TIMUR TERMINAL TIRTONADI KOTA SURAKARTA MENGGUNAKAN METODE BAYESIAN Rizka Nur Faizah; Sugito Sugito; Sudarno Sudarno
Jurnal Gaussian Vol 11, No 1 (2022): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/j.gauss.v11i1.33993

Abstract

The queuing system relates to customers and service facilities. Queuing theory designs service facilities to address service requests. Queues occur if the service capacity is not sufficient to provide services to many customers. The queuing phenomenon occurs on non-patas buses on the eastern route of Tirtonadi Terminal, Surakarta with Surabaya, Karanganyar, Wonogiri, Purwodadi and Pedesaan buses. The Bayesian method combines information from current research and previous studies with similar cases, and produces a posterior distribution to form a queuing system model and measure of service system performance. The bus queuing system model for Surabaya, Karanganyar, Wonogiri and Purwodadi has a Gamma-distributed arrival and service pattern. Pedesaan buses has an arrival pattern with a Gamma distribution and a service pattern with an Inverse Gamma distribution. Each line has 1 bus line as a service system, FIFO queue discipline, the number of customer capacity and call sources is not limited. The Surabaya buses has the highest probability of 93.49% that the line is idle and the Pedesaan buses  has the highest probability that the line will be busy serving at 89.50%. The queuing system are considered good because the five lines of service facilities are able to meet customer needs. Keywords: Tirtonadi Terminal, Bayesian, Posterior Distribution, Queue Models, System Performance Measures
PEMODELAN RETURN HARGA SAHAM MENGGUNAKAN MODEL INTERVENSI–ARCH/GARCH (Studi Kasus : Return Harga Saham PT Bayan Resources Tbk) Dea Manuella Widodo; Sudarno Sudarno; Abdul Hoyyi
Jurnal Gaussian Vol 7, No 2 (2018): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (512.454 KB) | DOI: 10.14710/j.gauss.v7i2.26642

Abstract

The intervention method is a time series model which could be used to model data with extreme fluctuation whether up or down. Stock price return tend to have extreme fluctuation which is caused by internal or external factors. There are two kinds of intervention function; a step function and a pulse function. A step function is used for a long-term intervention, while a pulse function is used for a short-term intervention. Modelling a time series data needs to satisfy the homoscedasticity assumptions (variance of residual is homogeneous).  In reality, stock price return has a high volatility, in other words it has a non-constant variance of residuals (heteroscedasticity). ARCH (Autoregressive Conditional Heteroscedasticity) or GARCH (Generalized Autoregressive Conditional Heteroscedasticity) can be used to model data with heteroscedasticity. The data used is stock price return from August 2008 until September 2018. From the stock price return data plot is found an extreme fluctuation in September 2017 (T=110) that is suspected as a pulse function. The best model uses the intervention pulse function is ARMA([1,4],0) (b=0, s=1, r=1). The intervention model has a non-constant variance or there is an ARCH effect. The best variance model obtained is ARMA([1,4],0)(b=0, s=1, r=1)–GARCH(1,1) with the AIC value is -205,75088. Keywords: Stock Return, Intervention, Heteroscedasticity, ARCH/GARCH