Claim Missing Document
Check
Articles

Found 32 Documents
Search

Statistical Inference for Modeling Neural Network in Multivariate Time Series Urwatul Wutsqa, Dhoriva; Subanar, Subanar; Guritno, Suryo; Soejoeti, Zanzawi
Jurnal ILMU DASAR Vol 9 No 1 (2008)
Publisher : Fakultas Matematika dan Ilmu Pengetahuan Alam Universitas Jember

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (136.781 KB)

Abstract

We present a statistical procedure based on hypothesis test to build neural networks model in multivariate time series case. The method involved strategies for specifying the number of hidden units and the input variables in the model using inference of R2 increment. We draw on forward approach starting from empty model to gain the optimal neural networks model. The empirical study was employed relied on simulation data to examine the effectiveness of inference procedure. The result showed that the statistical inference could be applied successfully for modeling neural networks in multivariate time series analysis.
PERAMALAN BEBAN LISTRIK DAERAH ISTIMEWA YOGYAKARTA DENGAN METODE SINGULAR SPECTRUM ANALYSIS (SSA) Utami, Herni; Sari, Yunita Wulan; Subanar, Subanar; Abdurakhman, Abdurakhman; Gunardi, Gunardi
MEDIA STATISTIKA Vol 12, No 2 (2019): 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 (740.583 KB) | DOI: 10.14710/medstat.12.2.214-225

Abstract

This paper will study forecasting model for electricity demand in Yogyakarta and forecast it for 2019 until 2024. Usually, electricity demand data contain seasonal. We propose Singular Spectral Analysis-Linear Recurrent Formula (SSA-LRF) method. The SSA process consists of decomposing a time series for signal extraction and then reconstructing a less noisy series which is used for forecasting. The SSA-LRF method will be used to forecast h-step ahead. In this study, we use monthly electricity demand in Yogyakarta for 11 year (2008 to 2018). The forecasting results indicates that the forecast using window length of L=26 have good performance with MAPE of 1.9%.
MODEL REPRESENTASI INFORMASI DAN PENGETAHUAN UNTUK PROYEK-PROYEK PERUSAHAAN DENGAN MENGGUNAKAN SEMANTIK ONTOLOGI Azhari, Azhari; Subanar, Subanar; Wardoyo, Retantyo; Hartati, Sri
JUTI: Jurnal Ilmiah Teknologi Informasi Vol 7, No 2, Juli 2008
Publisher : Department of Informatics, Institut Teknologi Sepuluh Nopember

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (341.632 KB) | DOI: 10.12962/j24068535.v7i2.a178

Abstract

This paper presents the utilization of knowledge management system for information and knowledge model development of enterprise projects. This information and knowledge management model is based on ontology semantic data model. The ontology data model is new technique for representing information and knowledge base on more semantically conception of meanings of objects, their properties, and relations between them that may arise within certain domain knowledge. The concern of the knowledge management model is to ensure that the model allows the process of creation, access, and utilization of data in a semantically manner (for querying process) and information or knowledge of enterprise projects. The experimentation shows that project ontology model has satisfied all consistent, valid, complete, and correct ontology model criteria and can be used for semantic reasoning computation. A prototype of the proposed model can access information and knowledge from the knowledge ontology model.   Kata Kunci: knowledge management system, semantic data model, ontology model, semantiq query, enterprise projects
Statistical Significance Test for Neural Network Classification Rezeki, Sri; Subanar, Subanar; Guritno, Suryo
Jurnal Natur Indonesia Vol 11, No 1 (2008)
Publisher : Lembaga Penelitian dan Pengabdian kepada Masyarakat Universitas Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (102.606 KB) | DOI: 10.31258/jnat.11.1.64-69

Abstract

Model selection in neural networks can be guided by statistical procedures, such as hypothesis tests, informationcriteria and cross validation. Taking a statistical perspective is especially important for nonparametric models likeneural networks, because the reason for applying them is the lack of knowledge about an adequate functionalform. Many researchers have developed model selection strategies for neural networks which are based onstatistical concepts. In this paper, we focused on the model evaluation by implementing statistical significancetest. We used Wald-test to evaluate the relevance of parameters in the networks for classification problem.Parameters with no significance influence on any of the network outputs have to be removed. In general, theresults show that Wald-test work properly to determine significance of each weight from the selected model. Anempirical study by using Iris data yields all parameters in the network are significance, except bias at the firstoutput neuron.
Asimtotik Model Multivariate Adaptive Regression Spline Otok, Bambang Widjanarko; Guritno, Suryo; Subanar, Subanar
Jurnal Natur Indonesia Vol 10, No 2 (2008)
Publisher : Lembaga Penelitian dan Pengabdian kepada Masyarakat Universitas Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (153.379 KB) | DOI: 10.31258/jnat.10.2.112-119

Abstract

Parameter estimation in MARS model executed by minimizing penalized least-squarer (PLS). Through somerequirement, asymtotic estimator characteristic from MARS prediction model has been successfully proven. Theresearch result shows that GCV can work properly to determine the best model that applied on MARS model. Solar’s vehicles produce opacity that exceed the standard limit of emition quality which was adjusted in Kepmen LH No.35 Year 1993, as large as 88 percent from 408 percent. Applying years, cylinder volume, type of machine, andvehicle’s radius are the variables that influences the opacity.
THE EFFECT OF DECOMPOSITION METHOD AS DATA PREPROCESSING ON NEURAL NETWORKS MODEL FOR FORECASTING TREND AND SEASONAL TIME SERIES Subanar, Subanar; Suhartono, Suhartono
Jurnal Teknik Industri Vol 8, No 2 (2006): DECEMBER 2006
Publisher : Institute of Research and Community Outreach - Petra Christian University

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (221.66 KB) | DOI: 10.9744/jti.8.2.pp. 156-164

Abstract

Recently, one of the central topics for the neural networks (NN) community is the issue of data preprocessing on the use of NN. In this paper, we will investigate this topic particularly on the effect of Decomposition method as data processing and the use of NN for modeling effectively time series with both trend and seasonal patterns. Limited empirical studies on seasonal time series forecasting with neural networks show that some find neural networks are able to model seasonality directly and prior deseasonalization is not necessary, and others conclude just the opposite. In this research, we study particularly on the effectiveness of data preprocessing, including detrending and deseasonalization by applying Decomposition method on NN modeling and forecasting performance. We use two kinds of data, simulation and real data. Simulation data are examined on multiplicative of trend and seasonality patterns. The results are compared to those obtained from the classical time series model. Our result shows that a combination of detrending and deseasonalization by applying Decomposition method is the effective data preprocessing on the use of NN for forecasting trend and seasonal time series.
Fuzzy C-Means Clustering Model Data Mining For Recognizing Stock Data Sampling Pattern Annatje Sumarauw, Sylvia Jane; Subanar, Subanar
IJCCS (Indonesian Journal of Computing and Cybernetics Systems) Vol 1, No 2 (2007): July
Publisher : IndoCEISS in colaboration with Universitas Gadjah Mada, Indonesia.

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22146/ijccs.2278

Abstract

AbstractCapital market has been beneficial to companies and investor. For investors, the capital market provides two economical advantages, namely deviden and capital gain, and a non-economical one that is a voting .} hare in Shareholders General Meeting. But, it can also penalize the share owners. In order to prevent them from the risk, the investors should predict the prospect of their companies. As a consequence of having an abstract commodity, the share quality will be determined by the validity of their company profile information. Any information of stock value fluctuation from Jakarta Stock Exchange can be a useful consideration and a good measurement for data analysis. In the context of preventing the shareholders from the risk, this research focuses on stock data sample category or stock data sample pattern by using Fuzzy c-Me, MS Clustering Model which providing any useful information jar the investors. lite research analyses stock data such as Individual Index, Volume and Amount on Property and Real Estate Emitter Group at Jakarta Stock Exchange from January 1 till December 31 of 204. 'he mining process follows Cross Industry Standard Process model for Data Mining (CRISP,. DM) in the form of circle with these steps: Business Understanding, Data Understanding, Data Preparation, Modelling, Evaluation and Deployment. At this modelling process, the Fuzzy c-Means Clustering Model will be applied. Data Mining Fuzzy c-Means Clustering Model can analyze stock data in a big database with many complex variables especially for finding the data sample pattern, and then building Fuzzy Inference System for stimulating inputs to be outputs that based on Fuzzy Logic by recognising the pattern.Keywords: Data Mining, AUz..:y c-Means Clustering Model, Pattern Recognition
A New Approach of Fuzzy-Wavelet Method’s Implementation in Time Series Analysis Hansun, Seng; Subanar, Subanar
IJCCS (Indonesian Journal of Computing and Cybernetics Systems) Vol 5, No 2 (2011): July
Publisher : IndoCEISS in colaboration with Universitas Gadjah Mada, Indonesia.

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22146/ijccs.2020

Abstract

      Abstract— Recently, many soft computing methods have been used and implemented in time series analysis. One of the methods is fuzzy hybrid model which has been designed and developed to improve the accuracy of time series prediction.      Popoola has developed a fuzzy hybrid model which using wavelet transformation as a pre-processing tool, and commonly known as fuzzy-wavelet method. In this thesis, a new approach of fuzzy-wavelet method has been introduced. If in Popoola’s fuzzy-wavelet, a fuzzy inference system is built for each decomposition data, then on the new approach only two fuzzy inference systems will be needed. By that way, the computation needed in time series analysis can be pressed.      The research is continued by making new software that can be used to analyze any given time series data based on the forecasting method applied. As a comparison there are three forecasting methods implemented on the software, i.e. fuzzy conventional method, Popoola’s fuzzy-wavelet, and the new approach of fuzzy-wavelet method. The software can be used in short-term forecasting (single-step forecast) and long-term forecasting. There are some limitation to the software, i.e. maximum data can be predicted is 300, maximum interval can be built is 7, and maximum transformation level can be used is 10. Furthermore, the accuracy and robustness of the proposed method will be compared to the other forecasting methods, so that can give us a brief description about the accuracy and robustness of the proposed method. Keywords—  fuzzy, wavelet, time series, soft computing
Stock Data Clustering of Food and Beverage Company Uyun, Shofwatul; Subanar, Subanar
IJCCS (Indonesian Journal of Computing and Cybernetics Systems) Vol 1, No 2 (2007): July
Publisher : IndoCEISS in colaboration with Universitas Gadjah Mada, Indonesia.

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22146/ijccs.2279

Abstract

AbstractCluster analysis can be defined as identifying groups of similar objects to discover distribution of patterns and interesting correlations in large data sets. Clustering analysis is important in the fields of pattern recognition and pattern classification. Over the years many methods have been developed for clustering data. In general, clustering methods can be categoried into two categories, i.e., fuzzy clustering and hard clustering. Fuzzy C-means is one of many methods of clustering based on fuzzy approach, while K-Means and K-Medoid are methods clustering based on crisp approach.This study aims to apply Fuzzy C-Means, K-Means and K-Medoid methods for clustering stock data in a jbod and beverage company. The main goal is to find a clustering method that can produce optimal clusters, The resulting clusters are validated using Dunn'• Index (DI). It is expected that the result of this reseach can be used to support decision making in the food and beverage company.Keywords : Clustering, Fuzzy C-Means, K-Means, K-Medoid, Cluster Validity, Dunn's Index (Dl)
Pemodelan Sistem Interaksi Obat dengan Menggunakan Fuzzy Inference System dan Pareto Optimality Yustina, Elena; Subanar, Subanar
IJCCS (Indonesian Journal of Computing and Cybernetics Systems) Vol 6, No 1 (2012): January
Publisher : IndoCEISS in colaboration with Universitas Gadjah Mada, Indonesia.

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22146/ijccs.2038

Abstract

Drug interactions is the interaction occurs between drugs that consumption simultaneously. Drug interactions can produce a good effect on patients, but not rarely produce adverse effects. Patients with diabetes and hypertension expected to control blood pressure and blood glucose levels to remain in normal circumstances, it is necessary to consider the use of medications for both diseases in order to produce effective therapies.Pareto optimality is a popular concept in the determination optimal solution of multiobjective problems. In determining the optimal solution of multiobjective problem should pay attention for each objective function, frequently conflicting objective functions. The interaction of two drugs has two objective function that is maximizing the positive effects and minimize negative effects. So its use is necessary to find optimal solutions to achieve the expected therapeutic. This research using Fuzzy Inference Sistem (FIS) to determine the appropriate medication to keep blood pressure and blood glucose levels of patients with hypertension and diabetes under control in normal and Pareto optimality to determine drug optimal solution.Fuzzy Inference System generates output choice of drug classes based on fuzzy rules in accordance with the patient's disease condition. Pareto optimality produces a pair solution for diabetes and hypertension drug that satisfy thresholds the minimum effective level (Minimum Effective Concentration; MEC) and maximum toxic levels (Minimum Toxic Concentration; MTC) of each drug.