Autoregressive Integrated Moving Average (ARIMA) is one of the linear model that good, flexible, and easy to use in univariate time series analysis forecasting. Some research activities forecasting also suggest Artificial Neural Network (ANN) as an alternative nonlinear for ARIMA has a good ability capture pattern while ANN pattern. models have been widely used prediction roll motion. can be combine hybrid take advantage modeling compared model. In this paper, we adapt methodology Deep (DNN)...

Maximum multivariate cumulative sum (Max-MCUSUM) is one of the single control charts that plot statistic as a representation mean vector and covariance matrix. The Max-MCUSUM has unknown specific distribution. objective this paper to propose bootstrap-based chart for which reference value predetermined in Phase I monitoring process. For various numbers quality characteristics correlation coefficients, limits estimated using bootstrap approach are presented paper. Furthermore, average run...

Credit is the main business of rural banks. distribution cannot be separated from risk default by debtor which has an impact on reducing credit quality. Worsening quality potential to reduce bank income because bank's comes loan interest income. Apart that, worsening also increasing burden provisions for losses productive assets. One effort that can made minimize predict so you identify early a decline in This research aims obtain significant features influence at Rural Banks and...

Statistical Process Control (SPC) has been widely used in industry and services. The SPC can be applied not only to monitor manufacture processes but also the Intrusion Detection System (IDS). In network monitoring intrusion detection, a powerful tool ensure system security stability network. Theoretically, Hotelling’s T2 chart detection. However, there are two reasons why is suitable used. First, detection data involves large volumes of high-dimensional process data. Second, requires fast...

Two types of control charts exist based on different quality characteristics: variable and attribute. These characteristics are commonly monitored using separate procedures. Only a few studies focused the utilization to monitor process with mixed characteristics. This study develops new concept chart Principal Component Analysis (PCA) Mix, that is PCA method can jointly handle continuous categorical data. The Kernel Density Estimation (KDE) used estimate limit. Through simulation studies,...

Views Icon Article contents Figures & tables Video Audio Supplementary Data Peer Review Share Twitter Facebook Reddit LinkedIn Tools Reprints and Permissions Cite Search Site Citation Suhartono, Muhammad Hisyam Lee, Dedy Dwi Prastyo; Two levels ARIMAX regression models for forecasting time series data with calendar variation effects. AIP Conf. Proc. 11 December 2015; 1691 (1): 050026. https://doi.org/10.1063/1.4937108 Download citation file: Ris (Zotero) Reference Manager EasyBib Bookends...

Views Icon Article contents Figures & tables Video Audio Supplementary Data Peer Review Share Twitter Facebook Reddit LinkedIn Tools Reprints and Permissions Cite Search Site Citation Shindi Shella May Wara, Dedy Dwi Prastyo, Heri Kuswanto; Value at risk estimation with hybrid-SVR-GARCH-KDE model for LQ45 portfolio optimization. AIP Conference Proceedings 27 January 2023; 2540 (1): 080013. https://doi.org/10.1063/5.0107539 Download citation file: Ris (Zotero) Reference Manager EasyBib...

Outliers presence may lead to misdetection on out-of-control observations in Phase II, therefore, they should be cleaned I. This paper proposes PCA Mix based T2 chart with Kernel Density control limit for mixed continuous and categorical data. Simulation studies are conducted evaluate the performance of proposed detecting outliers from clean contaminated The has better than benchmark monitoring For data, optimal situation when data generated multinomial distribution balanced parameters. is...

Maximum multivariate cumulative sum (Max-MCUSUM) is one of the single control charts proposed for joint monitoring mean and variability independent observation. Since many applications yield time series data, it important to develop Max-MCUSUM chart autocorrelated processes. In this paper, we propose a based on residual multioutput least square support vector regression (MLS-SVR). The optimal parameters MLS-SVR model are calculated using historical in-control data limit estimated bootstrap...

Monthly data about oil production at several drilling wells is an example of spatio-temporal data. The aim this research to propose nonlinear model, i.e. Feedforward Neural Network - Vector Autoregressive (FFNN-VAR) and FFNN Generalized Space-Time (FFNN-GSTAR), compare their forecast accuracy linear VAR GSTAR. These models are proposed applied for forecasting monthly three in East Java, Indonesia. There 60 observations that be divided two parts, the first 50 training last 10 testing results...

Water supply management effectively becomes challenging due to the human population and their needs have been growing rapidly. The aim of this research is propose hybrid methods based on Singular Spectrum Analysis (SSA) decomposition, Time Series Regression (TSR), Automatic Autoregressive Integrated Moving Average (ARIMA), known as SSA-TSR-ARIMA, for water demand forecasting. Monthly data frequently contain trend seasonal patterns. In research, two groups different were developed proposed,...

Short-term electricity load forecasting plays a vital role in determining the energy supplied. It has daily, weekly, and yearly harmonics patterns. In Muslim countries, it is influenced by religious celebrations held on Islamic Hijri calendar. This study aims to compare Dynamic Harmonic Regression its hybrid version involving calendar variation effect. designed two scenarios. Scenario 1 shows that out-of-sample can capture both seasonal patterns, while 2 performance only patterns exist. The...

Gamma distribution is a general type of statistical that can be applied in various fields, mainly when the data not symmetrical. When predictor variables also affect positive outcome, then gamma regression plays role. In many cases, give effect to several responses simultaneously. this article, we develop multivariate (MGR), which one non-linear with response follow (MG) distribution. This work provides parameter estimation procedure, test statistics, and hypothesis testing for significance...

Intermittent demand data is one of the types with a very random pattern, for example data. The will have value (not zero) if there demand. If no demand, zero. usually called customer or sales an item that not sold every time. general problem always continuous but intermittent. This natural fact makes intermittent easy to predict. Standard methods used predict are Croston. Single exponential smoothing (SES) also commonly in practice. Croston and generally produce static forecasts. study...

Abstract Multivariate Adaptive Regression Spline (MARS) is a nonparametric regression method that can accommodate additive effects and interaction between predictor variables. Generally, MARS has been used for modeling pairs of data with continuous or categorical responses. One type needs special attention in count data. The often encountered, especially the health sector. existence motivates development theory application method, which Poisson (MAPRS). MAPRS combination regression. It...