Khairil Anwar Notodiputro

A study of area clustering using factor analysis in small area estimation (An analysis of per capita expenditures of subdistricts level in regency and municipality of Bogor)

Empirical Best Linear Unbiased Prediction (EBLUP) is one of indirect estimating methods which used to estimate parameters of small areas. EBLUP methods works in using auxiliary variables of area while adding the area random effects. In estimating non-sampled area, the standard EBLUP can no longer be used due to no information of area random effects. To obtain more proper estimation methods for non sampled area, the standard EBLUP model has to be modified by adding cluster information. The aim of this research was to study clustering methods using factor analysis by means of simulation, provide better cluster information. The criteria used to evaluate the goodness of fit of the methods in the simulation study were the mean percentage of clustering accuracy. The results of the simulation study showed the use of factor analysis in clustering has increased the average percentage of accuracy particularly when using Ward method. The method was taken into account to estimate the per capita expenditures based on ...

Mengukur Indeks Kebahagiaan Mahasiswa IPB Menggunakan Analisis Faktor

Undergraduate students of Bogor Agricultural University are spread out in 9 Faculties and 1 School. The difference of faculties and schools illustrate the different characteristics and burdens of student lectures on each faculty and school. This distinction raises various assumptions about the level of student happiness in every faculty and school. Student happiness analysis is measured using loading factor obtained from Factor Analysis. Based on the analysis, found that Faculty of Animal Science is the happiest faculty with happiness index reaching 66.88 and the lowest index of happiness found in the Faculty of Human Ecology with happiness index of 62.39.

Pemodelan Loyalitas Konsumen Susu Pertumbuhan dalam Mengikuti Program Rewards Menggunakan Metode Random Forest dan Neural Network

Business competition in Indonesia has been becoming more competitive. This is the reason for companies to create a strategy in maintaining their customers through of loyalty program. However, on one of the rewards programs in nutrition companies, 37.62% of registered members have left the program or commonly known as churn. The classification model of customer loyalty is built to anticipate this, based on their profiles and activities in the program during the prediction period. The classification model is built for the two largest brands with random forest and neural network methods. These two methods are evaluated and compared based on the ROC (Relative Operating Characteristics) curve by considering the area under curve (AUC) value. The performance of these two methods are not significantly different, but neural network method yields greater AUC value, both in modeling of brand A and brand B.

A comparative study of approximation methods for maximum likelihood estimation in generalized linear mixed models (GLMM)

Maximum likelihood estimates in GLMM are often difficult to be obtained since the calculation involves high dimensional integrals. It is not easy to find analytical solutions for the integral so that the approximation approach is needed. In this paper, we discuss several approximation methods to solve high dimension integrals including the Laplace, Penalized Quasi likelihood (PQL) and Adaptive Gaussian Quadrature (AGQ) approximations. The performance of these methods was evaluated through simulation studies. The ‘true’ parameter in the simulation was set to be similar with parameter estimates obtained by analyzing a real data, particularly salamander data (McCullagh & Nelder, 1989). The simulation results showed that the Laplace approximation produced better estimates when compared to PQL and AGQ approximations in terms of their relative biases and mean square errors.

Small area estimation of per capita expenditures using robust empirical best linear unbiased prediction (REBLUP)

A small area is an area with small sample size to estimate parameters in survey sampling. The direct estimation will produce inaccurate estimation since the sample size is not enough to produce estimation with acceptable precision. Small Area Estimation (SAE) is a solution to obtain more precise estimation in a small area. A well-known method in SAE is an empirical best linear unbiased prediction (EBLUP). EBLUP is the estimator of small area means. It will provide an accurate estimation under normality assumptions but it can be sensitive when the data are contaminated by outliers. In this article, we discussed a resistant method in SAE, i.e. robust empirical best linear unbiased prediction (REBLUP). We apply REBLUP from unit-level models to the data obtained from the National Socio-economic Survey (SUSENAS). The means of per capita expenditures are calculated for all small areas. We compare the estimates of per capita expenditure in the small area using direct estimation, EBLUP and REBLUP methods using da...

A Comparative Study of Imputation Methods for Estimation of Missing Values of Per Capita Expenditure in Central Java

Missing values in repeated measurements have attracted concerns from researchers in the last few years. For many years, the standard statistical methods for repeated measurements have been developed assuming that the data was complete. The standard statistical methods cannot produce good estimates if the data suffered substantially by missing values. To overcome this problem the imputation methods could be used. This paper discusses three imputation methods namely the Yates method, expectation-maximization (EM) algorithm, and Markov Chain Monte Carlo (MCMC) method. These methods were used to estimate the missing values of per-capita expenditure data at sub-districts level in Central Java. The performance of these imputation methods is evaluated by comparing the mean square error (MSE) and mean absolute error (MAE) of the resulting estimates using linear mixed models. It is showed that MSE and MAE produced by the Yates method are lower than the MSE and MAE resulted from both the EM algorithm and the MCMC method. Therefore, the Yates method is recommended to impute the missing values of per capita expenditure at sub-district level.

Synthetic Over Sampling Methods for Handling Class Imbalanced Problems : A Review

Class imbalanced commonly found in any real cases. Class imbalanced occur if one of the classes has smaller amount, called minority class, than other class (majority class). The problem of imbalanced data is usually associated with misclassification problem where the minority class tends to be misclassified as compared to the majority class. There are two approaches should be performed to solve imbalanced data problems, those are solution at data level and solution at algorithm level. Over sampling approach is used more frequently than the other data level solution methods. This study gives review of synthethic over sampling methods for handling imbalance data problem. The implementation of different methods will produce different characteristics of the generated synthetic data and the implementation of appropriate methods must be adapted to the problems faced such as the level and pattern of imbalanced data of data available. Results of the review show that there is no absolute methods that are more efficient in dealing with the class imbalance. However, the class imbalance problem depends on complexity of the data, level of class imbalance, size of data and classifier involved. Determination of over sampling strategy will affect the outcome of the over sampling. So it is still open better development oversampling methods for handling the class imbalance. The selection classifier and evaluation measures are important to get the best results. Statistical test approach is needed to assess the theoritical propertis of synthetic data and evaluate missclassification in addition to the evaluation methods that have been used.

Small area estimation models with time factor effects for repeated measurement data

There has been an increasing demand for small area statistics based on sample survey in last few years. Generally, sample survey is designed for a specific area with large sample size. Since the sample size of a specific area is small, a direct estimation for sub population fails to provide enough precision. On the other hand, the per capita expenditure as a measure of prosperity and well-being is becoming an important issue in developing countries. In Indonesia, the available data related to this issue is usually measured repeatedly in quarterly using the national social-economic survey (Susenas). There are two problems to analyze Susenas data in district level. Firstly, the sample size in each quarter is relatively small. Secondly, the auxiliary data usually is available yearly only. To overcome this problem, a proper estimation technique was exercised using modeling systems. An extended of small area estimation (SAE) technique, based on both Fay-Herriot and Rao-Yu models in repeated measurement using Susenas data was discussed. Especially, SAE models with time factor effects motivated by Rao and Yu [7] were proposed. The performances of these models were evaluated by comparing the root mean square error (RMSE) and coefficient of variation (CV) of the estimators. The results showed that Rao-Yu with time factor effect models produced smaller RMSE which led to a significant reduction in a CV relative to other models.

Estimation of unemployment rates using small area estimation model by combining time series and cross-sectional data

Labor force surveys conducted over time by the rotating panel design have been carried out in many countries, including Indonesia. Labor force survey in Indonesia is regularly conducted by Statistics Indonesia (Badan Pusat Statistik-BPS) and has been known as the National Labor Force Survey (Sakernas). The main purpose of Sakernas is to obtain information about unemployment rates and its changes over time. Sakernas is a quarterly survey. The quarterly survey is designed only for estimating the parameters at the provincial level. The quarterly unemployment rate published by BPS (official statistics) is calculated based on only cross-sectional methods, despite the fact that the data is collected under rotating panel design. The study purpose to estimate a quarterly unemployment rate at the district level used small area estimation (SAE) model by combining time series and cross-sectional data. The study focused on the application and comparison between the Rao-Yu model and dynamic model in context estimating...